What is shown in the tables? Explanation of headings & data.

The content of four tables of database is explained in detail below: Tumors, ACF, Min, and Huge.

- [Tumor] table shows the following data

• 0- Numb. Each line has a number. Il helps to rank a new chemopreventive agent in comparison with all others. It's also used to fetch full data on a given agent in the "Huge" table.
• 1- Treatment = agent name & dose, and //carcinogen when it is not AOM or DMH, and /mice when it is not rats. Doses are reported as percent of the diet (w/w), or ppm (1000 ppm = 0.1%), or mg/kg (body weight). Most studies were done in rats injected with the carcinogens azoxymethane or dimethylhydrazine. When mice were used, or other carcinogens, it is reported in column 1.
  - ! - : Caution, potency cannot be accurately estimated when the number of tumor bearing rats in the control group is too small. This number is < 8 in studies tagged with a "!" : data may be imprecise.
  
• 2- Category or class of agents. "+" indicates a recent study, not reported in the original publication Corpet & Tache 2002. Categories: 1, polyethylene glycol (PEG); 2, non-steroidal anti-inflammatory drug (NSAID); 3, lipid or oil; 4, phytochemical; 5, fiber or bacteria; 6, calcium or salt; 7, retinoid, vitamin A or D derivative; 8, DFMO or anti-amine; 9, others.
• 3- Author and year (Reference of the study): A click on author's name links to PubMed abstract (click works in the database tables, not here).
• 4- Potency /tumor incidence: potency of each agent to reduce the incidence of intestinal adenoma & adenocarcinoma. Potency was estimated by the ratio of tumor incidence in control rats divided by tumor incidence in treated rats. Thus, potency tells the times-fold reduction in tumor incidence due to the agent. More explanations on potency.
• 5- Tumor percent inhibition: % reduction of colonic tumor incidence. Percent inhibition is equal to (100 - 100/potency). Very potent agents reduce tumor incidence by more than 80%. The average chemopreventive agent reduces the incidence by 50%.
• 6- Cancer percent inhibition: % reduction of colon cancer incidence. A negative value indicates a promotion: more rats with cancer in the treated group than in the control group.
• 7- Potency /multiplicity: potency to reduce the number of colonic tumors per rat. Potency was estimated by the ratio of mean tumor multplicity in control rats or mice divided by mean tumor multiplicity in treated rats or mice. More explanations on potency. Caution: some studies report the number of tumors per tumor-bearing rats, other report the number of tumors per rat (including rats without tumor).
• 8- Protocol or experimental design: the preventive agent was given to the animals either during the initiation with chemical carcinogen (init), or after the end of initiation (post), or during both periods (both).
• 8 bis- Blinded scoring: Each colon received a "blinding code" before being scored for tumors. "blind" = the observer was "blinded", "unaware", or "without knowing" rat treatment. Many authors do not state it explicitly (although they blind as a good laboratory practice).
• 9- Rodents' species, strain and gender: R, rats; M, mice; ', male; ", female. e.g., Rf' = Rats Fischer 334 male (see all abbreviations).
• 10- Carcinogen: the abbreviation of carcinogen is followed by the number of injections, and the dose of each injection (mg/kg BW).
• 11- Weeks: number of weeks between the first carcinogen injection and the end of the study (sacrifice).
• 12-16- Raw data: tumor incidences:
  col.12: actual numbers of tumor bearing rats in control group.
  col.13: total number of rats (all), in control group.
  col.14: actual numbers of tumor bearing rats in treated group.
  col.15: total number of rats in treated group.
  col.16: calculated potency = (12/13) / (14/15)
  Some study reports tumor incidence as percentages (indicated by %).
  Caution: potency cannot be accurately estimated when the number of tumor bearing rats in the control group is too small: numbers smaller than eight are tagged with a "!" and yield dubious data.
• 17-21- Raw data: Cancer incidence:
  col.17: actual numbers of rats with invasive cancer in control group.
  col.18: total number of rats (all), in control group.
  col.19: actual numbers of rats with invasive cancer in treated group.
  col.20: total number of rats in treated group.
  col.21: calculated % inhibition of cancer = 100 - 100* ((19/20) / (17/18))
  Some study reports cancer incidence as percentages (indicated by %).
• 22-24- Raw data: Tumor multiplicity:
  col.22: Actual numbers of tumors per rats, in control group
  col.23: Actual numbers of tumors per rats, in treated group
  col.24: calculated potency = 22 / 23
  Caution: some study report the number of tumors per tumor-bearing rats, some per rat (including rats without tumor)


• Ranking: Agents in the [Tumor] table are ranked by potency.
  The table can be sorted on other criteria (e.g., treatment). To change sorting, click on column heading (once: up, twice: down), or use the "select" box at table bottom.

- [ACF] table shows the following data:

• 0- Numb. Each line has a number. Il helps to rank a new chemopreventive agent in comparison with all others. It's also used to fetch full data on a given agent in the "Huge" table.
• 1- Treatment = agent name & dose, and //carcinogen when it is not AOM or DMH, and /mice when it is not rats. Doses are reported as percent of the diet (w/w), or ppm (1000 ppm = 0.1%), or mg/kg (body weight). Most studies were done in rats injected with the carcinogens azoxymethane or dimethylhydrazine. When mice were used, or other carcinogens, it is reported in column 1.
  - ! - : Caution, potency cannot be accurately estimated when the number of ACF in the control group is too small. This number is < 8 in studies tagged with a "!" : data may be imprecise.
  
  - * - : not ACF, but *MDF or *BCAC: 13 years after ACF discovery, other micro-lesions were described, MDF (Mucin Depleted Foci), and BCAC (Beta-Catenin-Accumulated Crypts). These putative preneoplastic lesions are discussed on a companion website with photos: [MDF - BCAC]. The few studies with MDF or BCAC endpoints are reported in the ACF table, tagged with *, *MDF or *BCAC. Data are reported as MDF or BCAC numbers instead of ACF numbers in those studies. "MDF - BCAC"
• 2- Category of agents. "+" indicates a recent study, not reported in the original publication Corpet & Tache 2002. Classes: 1, polyethylene glycol (PEG); 2, non-steroidal anti-inflammatory drug (NSAID); 3, lipid or oil; 4, phytochemical; 5, fiber or bacteria; 6, calcium or salt; 7, retinoid, vitamin A or D derivative; 8, DFMO or anti-amine; 9, others.
• 3- Author and year (Reference of the study): A click on author's name links to PubMed abstract (click works in the database tables, not here).
• 4- Potency /ACF: potency of each agent to reduce the number of aberrant crypt foci. Potency was estimated by the ratio of mean ACF number in control rats divided by mean ACF number in treated rats. Thus, potency tells the times-fold reduction in ACF number due to the treatment. More explanations on potency.
• 5- Percent inhibition: % reduction in ACF number. Percent inhibition is equal to (100 - 100/potency). Very potent agents reduce ACF number by more than 80%. The average chemopreventive agent reduces the ACF number by 50%.
• 6- Potency on large ACF: times-fold reduction in the number of large ACF per colon. In most articles, large ACF contains more than 3 crypts. Empty cell: data missing in original article. More explanations on potency.
• 7- Potency on ACF size: potency to reduce the number of crypt per ACF. This end-point may better correlate to tumor inhibition than the total number of ACF per colon. Empty cell: data missing in original article.
• 8- Potency on Aberrant Crypts: potency to reduce the number of Aberrant Crypts per colon = col.4 x col.7 Empty cell: data missing in original article.
• 9- Protocol or experimental design: the preventive agent was given to the animals either during the initiation with chemical carcinogen (init), or after the end of initiation (post), or during both periods (both).
• 9 bis- Blinded scoring: Each colon received a "blinding code" before being scored for ACF. "blind" = the observer was "blinded", "unaware", or "without knowing" rat treatment. Many authors do not state it explicitly (although they blind as a good laboratory practice).
• 10- Rodents' species, strain and gender: R, rats; M, mice; ', male; ", female. e.g., Rf' = Rats Fischer 334 male (see all abbreviations).
• 11- Carcinogen dose: the abbreviation of carcinogen is followed by the number of injections, and the dose of each injection (mg/kg BW).
• 12- Weeks: first value indicates the treatment start, relative to the carcinogen injection(s) (e.g., -1 means that chemopreventive treatment started one week before AOM first injection. +1 means the treatment started one week after the last carcinogen injection). Second value indicates the end of chemopreventive treatment relative to the last carcinogen injection. The treatment length may be calculated approximatly by substrating the first value from second value, and adding the number of weekly carcinogen injections.
• 13- ACF number in control rats: Raw data, mean number of aberrant crypt foci (ACF) in the colon of control rats, not given the chemopreventive treatment (raw data). Number of ACF was arbitrarily set to 100 when the article does not report it, but gives the % decrease (e.g., in proc.AACR abstracts). Low numbers of ACF (e.g., <10) relate to "ACF/square cm", or were induced by PhIP or IQ
• 14- ACF number in treated rats: Raw data, mean number of ACF in the colon of treated rats (raw data).
• 15- Number of large ACF in controls: Raw data, mean number of large ACF in the colon of control rats, not given the chemopreventive treatment. In most studies, ACF containing more than 3 crypts were counted as "large" (raw data). Number of large ACF was arbitrarily set to 20 when the article does not report it, but gives the % decrease (e.g., in proc.AACR abstracts)
• 16- Number of large ACF in treated: Raw data, mean number of large ACF in the colon of treated rats (raw data).
• 17- ACF size in controls: Raw data, mean number of crypts per ACF in the colon of control rats, not given the chemopreventive treatment. This number increases with time after initiation, and thus depends on column "weeks"
• 18- ACF size in treated: Raw data, mean number of crypts per ACF in the colon of treated rats.
• 19- Agent (3 letters) or comment or additional data
• Ranking: Agents in the [ACF] table are ranked by potency. The table can be sorted on other criteria (e.g., Treatment). To change sorting, click on column heading (once: up, twice: down), or use the "select" box at table bottom.

- [Min] mice table shows the following data

• 1- Treatment & mutation, if it is not Min Apc 850 mutation. Doses are reported in col. 16.
• 2- Category of agents. "+" indicates a recent study, not reported in the original publication Corpet & Pierre 2003. Classes: 1, polyethylene glycol (PEG); 2, non-steroidal anti-inflammatory drug (NSAID); 3, lipid or oil; 4, phytochemical; 5, fiber or bacteria; 6, calcium or salt; 7, retinoid, vitamin A or D derivative; 8, DFMO or anti-amine; 9, others.
• 3- Author (Reference of the study): A click on author's name links to PubMed abstract (click works in the database tables, not here).
• 4- Year (Reference of the study).
• 5- Potency of each agent to reduce the number of polyps in the whole gut of Min mice (both small and large bowel). Potency was estimated by the ratio of (mean polyp number in control mice) divided by (mean polyp number in treated rats). Thus, potency tells the times-fold reduction due to treatment, in the number of intestinal polyps. More explanations on potency.
• 6- N, number of pooled studies from one article which were used to calculate the reported mean value (e.g., if an article reports that 100ppm and 200ppm of sulindac reduce the number of polyps 9-fold and 11-fold respectively, the mean 10-fold is reported in the table, and N=2).
• 7- Number of polyps (adenomas) in the small intestine of control mice (mutant Min mice without treatment).
• 8- Number of polyps in the small intestine of treated mice.
• 9- Treatment effect in the small intestine, given as a percentage (small percentage = great protection): number of polyps in treated mice, percent of number in control mice.
• 10- P value: a star * indicates a significant effect of treatment in the small bowel (significant protection or promotion).
• 11- Number of polyps (adenomas) in the colon (large intestine) of control mice.
• 12- Number of polyps in the colon of treated mice.
• 13- Treatment effect in the colon, given as a percentage: number of polyps in treated mice, percent of number in control mice (small percentage = great protection; 100=no effect; large values = big promotion).
• 14- P value: a star * indicates a significant effect of treatment in the colon (significant protection or promotion).
• 15- Specificity for the colon: ratio of small vs. large bowel effect (col.15 = col.9/col.13). A high value denotes an agent that inhibits colonic polyps more than small intestinal polyps.
• 16- Treatment: Doses are reported as percent of the diet (w/w), or ppm (1000 ppm = 0.1%), or mg/kg (body weight).
• 17- Duration: Length of treatment given in weeks (w) or days (d). A / denotes the treatment start, e.g., /d-21 treatment started 21 days before birth; /w3 treatment started at 3 weeks of age.
• Ranking: Agents in the [Min] mice table are ranked by potency. The table can be sorted on other criteria, e.g., category or name of agents. To change sorting, click on column heading (one click: 9-to-1, two clicks: 1-to-9), or use the "select" box at table bottom.

- The [Huge] table documentation was moved to Doc-Huge