jeancarloszambrano / seer_solid_tumor

R code for SEER data analysis of solid tumor in different populations

SEER (solid tumor)

R code for SEER data analysis of solid tumor in different populations.

SEER*Stat data + R + R Markdown

Goal

Create reproducible report for different cancer types and parameters with minimum modifications on user’s side to observe patterns of different cancer types in populations.

Data source

https://seer.cancer.gov/data/access.html

The Surveillance, Epidemiology, and End Results (SEER) Program of the National Cancer Institute (NCI) collects and publishes cancer data through a coordinated system of strategically placed cancer registries which cover near 30% of the USA population. Currently there are 18 SEER registries in the USA.

Parameters in the code

Parameters are useful when you want to re-render the same report with distinct values for various key inputs.

• Years of diagnoses (2004+);
• Age groups at diagnosis;
• Cancer type (defined by ICD-O-3 codes together with Histology codes);
• Which SEER registry to compare with other 17 SEER registries;
• How to describe the extent of disease progression in cancer patients (by American Joint Committee on Cancer (AJCC) staging system or Grade system);
• Which edition of AJCC staging to apply (6th or combination of 6th and 7th);

Some other parameters are also available to be edited, but they are secondary.

Statistics in the report

• Characteristics of study participants
• Crude incidence rates by strata
• Age-adjusted incidence rates by strata
• Kaplan-Meier survival curves
• Cox proportional hazards models

Directions

1. Download code and Create this folders' structure:

|   Project_SEER_solid_tumor.Rproj  - always start a new R session by opening this file first
|   
+---code
|   +---generic
|   |       analyze.R
|   |       functions.R
|   |       map_SEER.R
|   |       map_template.R
|   |       subset.R
|   |       
|   \---specific
|       \---Colorectal
|               main.R              - main file, which controls all other files
|               report_docx_html.Rmd
|               styles_docx.docx
|               
\---data
    +---case_listing
    |       caselisting_data.dic    - download from SEER*Stat prior running the code and place under this name
    |       caselisting_data.txt    - download from SEER*Stat prior running the code and place under this name
    |       
    +---intermediate
    \---public
        \---data_geo
                mapcoord.RData

Additional subfolders and files should be automatically generated after running code in 'main.R' file:

|   Project_SEER_solid_tumor.Rproj
|   
+---code
|   +---generic
|   |       analyze.R
|   |       functions.R
|   |       map_SEER.R
|   |       map_template.R
|   |       subset.R
|   |       
|   \---specific
|       \---Colorectal
|               main.R
|               report_docx_html.docx  - automatically generated Word style report
|               report_docx_html.html  - automatically generated HTML style report
|               report_docx_html.Rmd
|               styles_docx.docx
|               
+---data
|   +---case_listing
|   |   |   caselisting_data.dic
|   |   |   caselisting_data.txt
|   |   |   
|   |   \---Colorectal
|   |           reg_inc.csv
|   |           reg_surv.csv
|   |           us_inc.csv
|   |           us_surv.csv
|   |           
|   +---intermediate
|   |   \---Colorectal
|   |           flow_charts.Rdata
|   |           icdo3_recode.Rdata
|   |           popdata.Rdata
|   |           pop_std.Rdata
|   |           results_of_analyze.RData
|   |           table1.Rdata
|   |           
|   \---public
|       |   stdpop.19ages.txt
|       |   us.1990_2015.19ages.adjusted.txt
|       |   
|       \---data_geo
|               mapcoord.RData
|               
\---figures
    |   SEER_covered_areas.png
    |   
    \---Colorectal
            many PNG files

2. Download data from SEER Program

2.1. Request access https://seer.cancer.gov/data/access.html;

2.2. Install SEER*Stat when access is granted;

2.3. Download all records from SEER*Stat’s 'Case Listing Session' by following rules:

I. In Data Tab:

Select "SEER*Stat Database: Incidence - SEER 18 Regs Research Data + Hurricane Katrina Impacted Louisiana Cases, Nov 2016 Sub (1973-2014 varying)" Database Name (or the most recent release).

II. In Selection Tab of SEER*Stat:

a) uncheck all boxes

b) {Race, Sex, Year Dx, Registry, County.Year of diagnosis} = '2004','2005','2006','2007','2008','2009','2010','2011','2012','2013','2014' (or later years)

Starting from 2004 6th and 7th editions of AJCC staging became available, so this code works for 2004+ cases.

III. In Table Tab of SEER*Stat pre-select columns to speed up download of the data

Column:	Age recode with <1 year olds
	Race recode (White, Black, Other)
	Sex
	Year of diagnosis
	SEER registry
	County
	State-county
	CHSDA 2012
	State
	Site recode ICD-O-3/WHO 2008
	Behavior recode for analysis
	Primary Site - labeled
	Primary Site
	Histologic Type ICD-O-3
	Behavior code ICD-O-3
	Grade
	Diagnostic Confirmation
	ICD-O-3 Hist/behav
	ICD-O-3 Hist/behav, malignant
	Histology recode - broad groupings
	Histology recode - Brain groupings
	Derived AJCC Stage Group, 7th ed (2010+)
	Derived AJCC Stage Group, 6th ed (2004+)
	Breast - Adjusted AJCC 6th Stage (1988+)
	Derived AJCC - Flag (2004+)
	Summary stage 2000 (1998+)
	SEER historic stage A
	COD to site recode
	SEER cause-specific death classification
	SEER other cause of death classification
	Survival months
	Survival months flag
	COD to site rec KM
	Vital status recode (study cutoff used)
	Type of follow-up expected
	Sequence number
	First malignant primary indicator
	Age recode with single ages and 85+
	Race recode (W, B, AI, API)
	Origin recode NHIA (Hispanic, Non-Hisp)
	Age at diagnosis
	Year of birth
	Month of diagnosis
	Month of diagnosis recode
	Patient ID
	Type of Reporting Source
	Insurance Recode (2007+)
	Marital status at diagnosis	

IV. Execute session.

V. Matrix -> Export -> Results as Text file -> save (.txt & .dic files fill be generated) under names 'caselisting_data.txt' and 'caselisting_data.dic' in ./data/case_listing/.

3. Run code in 'main.R' file for Colorectal cases

Always open R Project 'Project_SEER_solid_tumor.Rproj' in a new R session first, then run 'main.R'. It should be enough to generate reports automatically.

Projects allow to use relative path to the files in the code. As no one else will have exactly the same directory configuration as you. [http://r4ds.had.co.nz/]

In case of any issues here are directions on how to run code line-by-line (see sequence of files and lines):

-sequence-  -file-     -lines-     -comment-
1.          main.R      1-179
2.          analyze.R   1-47
3.          subset.R    All         this is the most time consuming part
4.          analyze.R   55-1002
5.          map_SEER.R  All
6.          analyze.R   1008-1016
7.          main.R      185-209

4. Run code with different parameters

4.1. Copy folder "Colorectal" from ./code/spacific and pate it to the same directory, rename new folder. Name has to be the same as 'cancer_type' variable in 'main.R' file.

4.2. In a new folder modify parameters in file 'main.R' and just 1 line in code-chunk called "change" in file 'report_docx_html.Rmd'. Read inline comments in 'main.R' for more instructions.

4.3. Run 'main.R' file or as described above line-by-line in order to debug.

Notes for SEER data analysis:

1. Use different data subsets for incidence rates and survival analyses;

Step	Selection Criteria	Incidence Rates	Survival Analyses
1	All cancer diagnoses in SEER registries for selected age groups, 2004 - 2014	o	o
2	Colorectal Cancer	o	o
3	Malignant	o	o
4	Microscopically confirmed	x	o
5	First primary: 'One primary only' or '1st of 2 or more primaries'	x	o
6	Active follow-up	x	o
7*	Reporting Source NOT 'Autopsy only' and NOT 'Death certificate only'	x	o

where 'x' means that we did not use these criteria, 'o' - applied these criteria;

'*' means that step 6 and 7 are usually do the same, but in rare cases might mean different things.

2. Read code comments.

50 states map

The map can be reproduced without access to SEER data files. There is templare R file and an example file which put SEER covered counties on the map using ggplot. You can actually put any data on the county or state map.

To do this you will need to download only 3 files: 'map_SEER.R' (from this GitHub repository), 'mapcoord.RData' (from this GitHub repository), and 'us.1990_2015.19ages.adjusted.txt' (see a link shown in comments in 'map_SEER.R'). Don't forget to modify file pathes in 'map_SEER.R'. Here is a generated map:

Motivation:

1. usually proposed R packages work for 48 states (no Alaska, no Puerto Rico);

2. proposed packages do not show states' names;

3. usually those packages do work for states, but not county levels data;

4. wanted to make maps from original shape files to learn the process.

Session Information

R version 3.4.2 (2017-09-28)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 7 x64 (build 7601) Service Pack 1

Matrix products: default

locale:
[1] LC_COLLATE=English_United States.1252 
[2] LC_CTYPE=English_United States.1252   
[3] LC_MONETARY=English_United States.1252
[4] LC_NUMERIC=C                          
[5] LC_TIME=English_United States.1252    

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
 [1] bindrcpp_0.2      R.utils_2.5.0     R.oo_1.21.0      
 [4] R.methodsS3_1.7.1 knitr_1.17        rmarkdown_1.6    
 [7] rgdal_1.2-12      rgeos_0.3-25      mapproj_1.2-5    
[10] maps_3.2.0        maptools_0.9-2    sp_1.2-5         
[13] plotrix_3.6-6     gridExtra_2.3     forcats_0.2.0    
[16] stringr_1.2.0     scales_0.5.0      survival_2.41-3  
[19] SEER2R_1.0        dplyr_0.7.4       purrr_0.2.3      
[22] readr_1.1.1       tidyr_0.7.1       tibble_1.3.4     
[25] ggplot2_2.2.1     tidyverse_1.1.1   plyr_1.8.4       

loaded via a namespace (and not attached):
 [1] Rcpp_0.12.13     lubridate_1.6.0  lattice_0.20-35  png_0.1-7       
 [5] assertthat_0.2.0 rprojroot_1.2    digest_0.6.12    psych_1.7.8     
 [9] R6_2.2.2         cellranger_1.1.0 backports_1.1.1  evaluate_0.10.1 
[13] highr_0.6        httr_1.3.1       rlang_0.1.2      lazyeval_0.2.0  
[17] readxl_1.0.0     Matrix_1.2-11    labeling_0.3     splines_3.4.2   
[21] foreign_0.8-69   munsell_0.4.3    broom_0.4.2      compiler_3.4.2  
[25] modelr_0.1.1     pkgconfig_2.0.1  mnormt_1.5-5     htmltools_0.3.6 
[29] tidyselect_0.2.0 grid_3.4.2       nlme_3.1-131     jsonlite_1.5    
[33] gtable_0.2.0     magrittr_1.5     stringi_1.1.5    reshape2_1.4.2  
[37] xml2_1.1.1       tools_3.4.2      glue_1.1.1       hms_0.3         
[41] yaml_2.1.14      parallel_3.4.2   colorspace_1.3-2 rvest_0.3.2     
[45] bindr_0.1        haven_1.1.0     

References

1. Stat software citation: Surveillance Research Program, National Cancer Institute SEER*Stat software (seer.cancer.gov/seerstat) version 8.3.4.

2. SEER data citation: Database name=Incidence - SEER 18 Regs Research Data + Hurricane Katrina Impacted Louisiana Cases, Nov 2016 Sub (1973-2014 varying) - Linked To County Attributes - Total U.S., 1969-2015 Counties.

3. R Core Team (2017). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL https://www.R-project.org/.

4. RStudio Team (2016). RStudio: Integrated Development for R. RStudio, Inc., Boston, MA URL http://www.rstudio.com/.

P.S.

Great books:

1. R in Action: Data Analysis and Graphics with R 2nd Edition by Robert Kabacoff;
2. http://r4ds.had.co.nz/ R for Data Science by Hadley Wickham and Garrett Grolemund;
3. Dynamic Documents with R and knitr by Yihui Xie.