Defining response to radiotherapy in rectal cancer using magnetic resonance imaging and histopathological scales

Table 1 Characteristics of studies reporting on good or poor response based upon histopathology

Ref.	Year	Chemotherapy protocol with radiotherapy	Radiotherapy protocol (Gy)	Surgical procedures	TME	Time to surgery (wk)	Cancer stage pre neo-adjuvant therapy	Adjuvant therapy
Gambacorta et al[21]	2004	Ralitrexed	50.4	APR/AR/Col-Anal resection/Stoma	Y	6-8	Stage 2 or 3	Y
Pucciarelli et al[28]	2004	Fluorouracil, leucovorin carboplatin, oxaliplatin	45-50.4	APR/AR/Hartmann’s	Y	2-8	T2/3/4, N0/1/2	Y
Beddy et al[17]	2008	Fluorouracil	45-50	APR/AR	Y		T3/4, N1/2
Giralt et al[22]	2008	Tegafir uracil, leucovorin	45 + 9 boost	APR/AR	Y	4-6	T3/4, N0/1/2	Y
Horisberger et al[24]	2008	Capecitabine, irinotecan	50.4	APR/AR/stoma	Y	4-7	T2/3/4, N+
Suárez et al[31]	2008	Fluoropyridine-based	50.4	APR/AR/Hartmann’s	Y	6	Stage 2 or 3	Y
Bujko et al[18]	2010	Fluorouracil, leucovorin	50.4	APR/AR/Hartmann’s	Y	4-6	Stage 2 or 3	Y
Avallone et al[13]	2011	Fluorouracil, levo-folinic acid, ralitrexed, oxaliplatin	45.0	APR/AR/Stoma	Y	< 8	T3/4, N0/1/2	Y
Eich et al[19]	2011	Fluorouracil	50.4	APR/AR/TEMS/Intersphincteric Surgery	Y	4-6	Stage 1,2 or 3	Y
Min et al[27]	2011	Fluorouracil, leucovorin	50.4	APR/AR	Y	6	T3/4, N0/1/2
Shin et al[30]	2011	Fluorouracil	25-50.4	APR/AR/Pan		4-6	T3/4
Huebner et al[25]	2012	Fluorouracil		APR/AR			T1/2/3/4, N0/1/2	Y
Lim et al[26]	2012	Capecitabine, fluorouracil, leucovorin	44-46+4.6 boost	Radical Proctectomy	Y		T3/4, N+	Y
Roy et al[29]	2012	Capecitabine, fluorouracil	45-50		Y	4-6	T1/2/3/4, N0/1/2	Y
Vallböhmer et al[32]	2012	Fluorouracil	50.4	APR/AR	Y		T3/4, N0/1/2
Winkler et al[33]	2012	Capecitabine, oxaliplatin	45-50.4		Y	4-6	Stage 2 or 3	Y
Elezkurtaj et al[20]	2013	Fluorouracil	50.4		Y	4-6
Hermanek et al[23]	2013			APR/AR/Hartmann’s	Y			Y
Fokas et al[14]	2014	Fluorouracil	50.4	APR/AR	Y	4-6	T3/4 or any T and N+	Y
Santos et al[16]	2014	Fluorouracil	50.4	APR/AR	Y	< 8	T2N+ or T3/4	Y
Hav et al[15]	2015	Fluorouracil, cetuximab, oxaliplatin	25-45	AR/Hartmann’s	Y	6-8	T3/4 or any T and N+

APR: Abdominoperineal resection; AR: Anterior resection; Pan: Panproctocolectomy; Col-Anal: Colorectal and anal resection; TME: Total mesorectal excision; Gy: Gray.

Table 2 Summary of histopathological tumour regression grade scales available in the literature for rectal cancer after neo-adjuvant treatment

TRG scale	Mandard
	(Low no. - More regression)[43]
0
1	Complete regression - absence of residual cancer and fibrosis
2	Presence of rare residual cancer
3	An increase in the number of residual cancer cells, but predominantly fibrosis
4	Residual cancer outgrowing fibrosis
5	Absence of regressive changes
TRG scale	Modified Mandard (Ryan)
	(Low no. - More regression)[37]
0
1	TRG 1 and 2 of the Mandard scale
2	TRG 3 of the Mandard scale
3	TRG 4 and 5 of the Mandard scale
4
5
TRG scale	Werner and Hoffler
	(Low no. - More regression)[41]
0
1	0% viable tumour cells
2	< 10% viable tumour cells
3	10%-50% viable tumour cells
4	> 50% viable tumour cells
5	No regression
TRG scale	Dworak
	(Low no. - Less regression)[35]
0	No regression
1	Dominant tumour mass with obvious fibrosis and/or vasculopathy
2	Dominant fibrotic change with few
	tumour cells or groups(easy to find)
3	Very few tumour cells in fibrotic tissue with or without mucous substance
4	No tumour cells, only fibrotic mass (total regression or response)
5
TRG scale	Modified Dworak
	(Low no. - Less regression)[38]
0	No regression
1	Regression ≤ 25% of tumour mass (dominant tumour mass with obvious fibrosis and/or vasculopathy)
2	Regression > 25%-50% of tumour mass (dominantly fibrotic changes with few tumour cells of groups, easy to find)
3	Regression > 50% of tumour mass (very few tumour cells in fibrotic tissue with or without mucous substance)
4	Complete (total) regression (or response): no vital tumour cells
5
TRG scale	AJCC 7th Edition[48]
0	Complete-no viable cells present
1	Moderate-single cells/small groups of cancer cells
2	Minimal-residual cancer outgrown by fibrosis
3	Poor-minimal or no tumour kill, extensive residual cancer
4
5
TRG scale	Memorial Sloan-Kettering (Low no. - Less regression)[47]
0	0%-85% regression
1	86-99% regression
2	100% regression
3
4
5
TRG scale	Cologne
	(Low no. - Less regression)[40]
0
1	> 50 % Viable rectal tumour cells
2	10%-50% Viable rectal tumour cells
3	Near complete regression with < 10% Viable rectal tumour cells
4	Complete regression (pathologic complete remission and ypT0)
TRG scale	Bujko/Glynne Jones
	(Low no. - More regression)[18,44]
0	No cancer cells
1	A few cancer foci in less than 10% of tumour mass
2	Cancer seen in 10%-50% of tumour mass
3	Cancer cells seen in more than 50% of tumour mass
4
TRG scale	College of American Pathologists[50]
0	Complete response: No residual tumour
1	Marked response: Minimal residual cancer
2	Moderate response: Residual cancer outgrown by fibrosis
3	Poor or no response: Minimal or no tumour kill; extensive residual cancer
4
TRG scale	RCPath system
	(Low no. - More regression)[42]
0
1	No residual cells and/or mucus lakes only
2	Minimal residual tumour i.e., microscopic residual tumour foci only
3	No marked regression
4
TRG scale	RCRG system
	(Low no. - More regression)[34]
0
1	Sterilisation or only microscopic foci of adenocarcinoma with marked fibrosis
2	Marked fibrosis but macroscopic disease present
3	Little or no fibrosis with abundant macroscopic disease
4
TRG scale	Mod RCRG system
	(Low no. - More regression)[45]
0
1	Macroscopic features may be varied. Microscopy reveals no tumour or < 5% of area of abnormality
2	Macroscopic features may be varied. Microscopy reveals combination of viable tumour and fibrosis. Tumour comprises 5%-50% of overall area of abnormality
3	Macroscopic or microscopic features may not be significantly different. Over 50% comprises tumour. Some fibrosis may be present but no more than untreated cases
4
TRG scale	Japanese
	(Low no. - Less regression)[25]
0	No regression
1a	Minimal effect (necrosis less than 1/3)
1b	Mild effect (necrosis less than 2/3 but more than 1/3)
2	Moderate effect (necrosis more than 2/3 of the lesion)
3	No tumour cells
TRG scale	Ruo
	(Low no. - Less regression)[39]
0	No evidence of response
1	1% to 33% response
2	34% to 66% response
3a	67% to 95% response
3b	96% to 99% response
4	100% response (no viable tumour identified)
TRG scale	Junker and Muller
	(Low no. - Less regression)[46]
1	No regression
2a	> 10% residual tumour cells
2b	< 10% residual tumour cells
3	Total regression (no viable tumour cells)
TRG scale	Rodel
	(Low no. - Less regression)[36]
Poor	TRG 1 and 0 of the Dworak scale
Intermediate	TRG 2 and 3 of the Dworak scale
Complete	TRG 4 of the Dworak scale
TRG scale	Four point scale
	Swellengrebel et al[49]
pCR	Pathological complete response without residual primary tumour
Near pCR	Isolated residual tumour cells/small groups of residual tumour cells
Response	Stromal fibrosis outgrowing tumour
No response	No regression or those with stromal fibrosis outgrown by tumour
TRG scale	Modified Mandard TRGN by Dhadda et al[51]
TRGN 1	Complete regression with absence of residual cancer and fibrosis extending through the wall
TRGN 2	Presence of rare residual cancer cells scattered through the fibrosis
TRGN 3	An increased number of residual cancer cells, but fibrosis is still predominant

Table 3 Permutations of regression scales to define poor and good response

Poor response		Good response
TRG grading system	Studies that used the scale	TRG grading system	Studies that used the scale
Mandard TRG 3,4,5	Suárez et al[31]	Mandard TRG 1,2	Suárez et al[31]
	Santos et al[16]		Gambacorta et al[21]
			Santos et al[16]
Mandard TRG 4	Gambacorta et al[21]	Mandard TRG 2,3	Avallone et al[13]
	Giralt et al[22]
Mandard TRG 4,5	Avallone et al[13]	Mandard TRG 1,2,3	Roy et al[29]
	Roy et al[29]		Pucciarelli et al[28]
	Pucciarelli et al[28]		Shin et al[30]
	Shin et al[30]
Dworak 1	Winkler et al[33]	Dworak TRG 2,3,4	Huebner et al[25]
			Roy et al[29]
Dworak TRG 0,1	Huebner et al[25]	Dworak TRG 2,3	Fokas et al[14]
	Roy et al[29]
	Fokas et al[14]
Dworak TRG 1,2	Lim et al[26]	Dworak TRG 3,4	Lim et al[26]
			Elezkurtaj et al[20]
			Santos et al[16]
			Hav et al[15]
Dworak TRG 0,1,2	Elezkurtaj et al[20]	Dworak TRG 3	Winkler et al[33]
	Hav et al[15]
	Santos et al[16]
Rodel TRG 3 [Dworak 0,1]	Min et al[27]	Japanese TRG 2,3	Horisberger et al[24]
Rodel TRG 3 [Wittekind (mod Dworak 0,1)]	Hermanek et al[23]	Japanese TRG 3	Vallböhmer et al[32]
Japanese TRG 0,1a,1b	Horisberger et al[24]	Miller Junker TRG 2a and 2b	Vallböhmer et al[32]
Japanese TRG 1	Vallböhmer et al[32]	Cologne TRG 3 and 4	Vallböhmer et al[32]
Miller Junker TRG 1	Vallböhmer et al[32]	Glynne Jones TRG 1	Bujko et al[18]
Miller Junker TRG 1,2a	Eich et al[19]
Cologne TRG 1,2	Vallböhmer et al[32]
Glynne Jones TRG 3	Bujko et al[18]
Wheeler RCRG TRG 2	Beddy et al[17]

Table 4 Proportion of poor responders in the literature according to regression grades

TRG grading system	No. of reports (total 25 reports from 21 studies)	Proportion of poor responders	Lower limit of confidence Interval	Upper limit of confidence Interval
Mandard	8	34.9	22.8	49.4
Dworak	8	47.4	32.5	62.7
Junker/Muller	2	50.8	28.8	72.5
Japanese	2	35.0	20.4	52.9
Wheeler	1	38.9	30.8	47.7
Bujko/Glynne-Jones	1	22.1	15.8	30.0
Rodel based on Dworak	1	52.2	44.9	59.5
Rodel based on Wittekind (modified Dworak)	1	14.7	10.6	19.9
Cologne	1	7.1	3.2	14.8

Table 5 Study definitions of poor response according to histopathological tumour regression grade scales

Ref.	Year	TRG scale used (original disease application)	Are the scales reported accurately?	Poor response definition	Total (n)	Poor responders (n)	Average F/up in months	LR (%) 5 yr	DR (%) 5 yr	DFS (%)	OS (%)
Gambacorta et al[21]	2004	Mandard (oesophagus)	Yes	TRG 4	54	10	25
Pucciarelli et al[28]	2004	Mandard (oesophagus)	Yes	TRG 4 and 5	106	52	42
Beddy et al[17]	2008	Wheeler (rectal)	Yes	TRG 2	126	49	37	21		Yr. 5: 71
Giralt et al[22]	2008	Mandard (oesophagus)	No	TRG 4	68	7
Horisberger et al[24]	2008	Japanese Society for Cancer of the Colon and Rectum (rectal)	Yes	TRG 0 and 1a and 1b	59	26
Suárez et al[31]	2008	Mandard (oesophagus)	Yes	TRG 3 and 4 and 5	119	83	33	3.41	14.31	Yr. 2: 83.6
										Yr. 3: 73.8
Bujko et al[18]	2010	Glynne Jones/Bujko (rectal)	Yes	TRG 3	131	29	48	26	47	Yr. 4: 47
Avallone et al[13]	2011	Mandard (oesophagus)	Yes	TRG 4 and 5	63	9	60			Yr. 5: Prob free of recurrence 562
Eich et al[19]	2011	Müller and Junker (lung)	Yes	TRG 1 and 2a	72	28	28			Yr. 2: 76 ± 14.8
Min et al[27]	2011	Rodel (rectal based on Dworak)	Yes	Categorised as poor according to Rodel and based on TRG 0 and 1 on Dworak scale	178	93	43	21	31
Shin et al[30]	2011	Mandard (oesophagus)	Yes	TRG 4 and 5	102	50	40.3			Yr. 3: 72.6
Huebner et al[25]	2012	Dworak (rectal)	Yes	TRG 0+1	237	61
Lim et al[26]	2012	Dworak (rectal)	Yes	TRG 1+2	581	357	61	9.5	27.2	Yr. 5: 63.6	Yr. 5: 71.3
Roy et al[29]	2012	Dworak (rectal)	Yes	TRG 0 and 1	75	42				Yr. 2: 68.9	Yr. 2: 92.6
Roy et al[29]	2012	Mandard (oesophagus)	Yes	TRG 4 and 5	75	24				Yr. 2: 60.3	Yr. 2: 87.3
Vallböhmer et al[32]	2012	Japanese Society for Cancer of the Colon and Rectum (rectal)	Yes	TRG 1	85	23
Vallböhmer et al[32]	2012	Junker Miller (lung)	Yes	TRG 1	85	6					DNE
Vallböhmer et al[32]	2012	Cologne (oesophageal)	Yes	TRG 1 and 2	85	53					DNE
Winkler et al[33]	2012	Dworak (rectal)	No	TRG 1	33	9					DNE
Elezkurtaj et al[20]	2013	Dworak (rectal)	Yes	TRG 0,1 and 2	102	68
Hermanek et al[23]	2013	Rodel (rectal based on Wittekind and Tannapfel (rectal based on Dworak)	Yes	Categorised as poor according to Rodel and based on TRG 0and1 on Wittekind and Tannapfel (a modified Dworak scale)	225	33	92	15.9	27.9	Yr. 5: 63.6	Yr. 5: 75.8
Fokas et al[14]	2014	Dworak (rectal)	Yes	TRG 0+1	386	90	132	Yr. 10: 3.6	Yr. 10: 39.6	Yr. 10: 63%
Santos et al[16]	2014	Dworak (rectal)	Yes	TRG 0,1 and 2	144	85	56	3.5	16.4	Yr. 5: 68.1	Yr. 5: 69.1
Santos et al[16]	2014	Mandard (oesophagus)	Yes	TRG 3 and 4 and 5	144	69	56	4.3	20.3	Yr. 5: 61.7	Yr. 5: 60.7
Hav et al[15]	2015	Dworak (rectal)	Yes	TRG 0,1 and 2	76	48	20			No specific data but no correlation with DFS

¹Overall rate for total follow-up time;

²Probability of being free from recurrence (DFS rate not given). LR: Local recurrence; DR: Distant recurrence.

Table 6 Study definitions of good response according to histopathological tumour regression grade scales

Ref.	Year	TRG scale used (original disease application)	Are the scales reported accurately?	Good response definition	Total (n)	Good responders (n)	Average F/up in months	LR (%) 5 yr	DR (%) 5 yr	DFS (%)	OS (%)
Gambacorta et al[21]	2004	Mandard (oesophagus)	Yes	TRG 1 and 2	54	24	25
Pucciarelli et al[28]	2004	Mandard (oesophagus)	Yes	TRG 1 and 2 and 3	104	52	42			DNE	DNE
Horisberger et al[24]	2008	Japanese Society for Cancer of the Colon and Rectum (rectal)	Yes	TRG 2 and 3	59	33
Suárez et al[31]	2008	Mandard (oesophagus)	Yes	TRG 1 and 2	119	36	33	0	0	DNE
Bujko et al[18]	2010	Glynne Jones/Bujko (rectal)	Yes	TRG 1	131	40	48	9	34	Yr. 4: 67
Avallone et al[13]	2011	Mandard (oesophagus)	Yes	TRG 2 and 3	63	20	60			Yr. 5: Prob free of recurrence > 90%
Shin et al[30]	2011	Mandard (oesophagus)	Yes	TRG 1 and 2 and 3	102	52	40.3			Yr. 3: 74.1
Huebner et al[25]	2012	Dworak (rectal)	Yes	TRG 2 and 3 and 4	237	176
Lim et al[26]	2012	Dworak (rectal)	Yes	TRG 3 and 4	581	224	61	1.3	11.6	Yr. 5: 86.7	Yr. 5: 88.2
Roy et al[29]	2012	Dworak (rectal)	Yes	TRG 2 and 3 and 4	75	33				Yr. 2: 91.7	Yr. 2: 89.2
Roy et al[29]	2012	Mandard (oesophagus)	Yes	TRG 1 and 2 and 3	75	51				Yr. 2: 86.1	Yr. 2: 92.2
Vallböhmer et al[32]	2012	Japanese Society for Cancer of the Colon and Rectum (rectal)	Yes	TRG 3	85	23					DNE
Vallböhmer et al[32]	2012	Junker Miller (lung)	Yes	TRG 2aand2b	85	65					DNE
Vallböhmer et al[32]	2012	Cologne (oesophageal)	Yes	TRG 3 and 4	85	26					DNE
Winkler et al[33]	2012	Dworak (rectal)	No	TRG 3	33	6
Elezkurtaj et al[20]	2013	Dworak (rectal)	Yes	TRG 3 and 4	102	34
Fokas et al[14]	2014	Dworak (rectal)	Yes	TRG 2 and 3	386	256	132	Yr. 10: 8.0	Yr. 10: 29.3	Yr. 10: 73.6%
Santos et al[16]	2014	Dworak (rectal)	Yes	TRG 3 and 4	144	54	56	1.8	11.1	Yr. 5: 78.4	Yr. 5: 77.4
Santos et al[16]	2014	Mandard (oesophagus)	Yes	TRG 1 and 2	144	70	56	1.4	8.6	Yr. 5: 81.7	Yr. 5: 79.4
Hav et al[15]	2015	Dworak (rectal)	Yes	TRG 3 and 4	76	28	20			No specific data but no correlation with DFS

Overall rate for total follow-up time. LR: Local recurrence; DR: Distant recurrence; DNE: Data given but not extractable; DFS: Disease-free survival.

Table 7 Comparison of outcomes between good and poor responders

Ref.	Year	Good response defn.	Poor response defn.	LR %		P < 0.05	DR %		P < 0.05	DFS %		P < 0.05	OS %		P < 0.05	DSS		P < 0.05	Conclusion
				GR	PR		GR	PR		GR	PR		GR	PR		GR	PR
Pucciarelli et al[28]	2004	TRG 1 and 2 and 3	TRG 4 and 5							Better in GR		No	Better in GR		No				Good responders have better, non-statistically significant outcomes for DFS and OS
Suárez et al[31]	2008	TRG 1 and 2	TRG 3 and 4 and 5	0	3.4	NC	0	14.3	NC	Better in GR		Yes				Better in GR		No	Good responders have better, statistically significant DFS but have better, non significant LR, DR and DSS
Bujko et al[18]	2010	TRG 1	TRG 3	9	26	No	34	47	No	67	47	No							Good responders have better, non-statistically significant outcomes for LR, DR and DFS
Avallone et al[13]	2011	TRG 2 and 3	TRG 4 and 5							Prob > 90%	Prob 56%	Yes							Good responders have better, statistically significant DFS
Shin et al[30]	2011	TRG 1 and 2 and 3	TRG 4 and 5							74.1	72.6	No							Good responders have better, non-statistically significant outcomes for DFS
Lim et al[26]	2012	TRG 3 and 4	TRG 1 and 2	1.3	9.5	Yes	11.6	27.2	Yes	86.7	63.6	Yes	88.2	71.3	Yes				Good responders have better, statistically significant outcomes for LR, DR, DFS and OS
Roy et al[29]	2012	TRG 1 and 2 and 3	TRG 4 and 5							86.1	60.3	Yes	92.2	87.3	No				Good responders have better, statistically significant DFS but have better, non significant OS
Roy et al[29]	2012	TRG 2 and 3 and 4	TRG 0 and 1							91.7	68.9	No	89.2	92.6	No				Good responders had better, non-statistically significant outcomes for DFS. Good responders had poorer, non-statistically significant outcomes for OS
Vallböhmer et al[32]	2012	TRG 3	TRG 1										Better in GR		No				Good responders have better, non-statistically significant outcomes for OS
Vallböhmer et al[32]	2012	TRG 2a and 2b	TRG 1										Better in GR		No				Good responders have better, non-statistically significant outcomes for OS
Vallböhmer et al[32]	2012	TRG 3 and 4	TRG 1 and 2										Better in GR		No				There was no statistically significant difference for OS between good and poor responders
Fokas et al[14]	2014	TRG 2 and 3	TRG 0 and 1	8	3.6	No	29.3	39.6	Yes	73.6	63	Yes							Good responders have better, statistically significant outcomes for DR and DFS. Good responders had poorer, non-statistically significant outcomes for LR
Santos et al[16]	2014	TRG 1 and 2	TRG 3 and 4 and 5	1.4	4.3	NC	8.6	20.3	NC	81.7	61.7	Yes	79.4	60.7	Yes				Good responders have better, statistically significant outcomes for DFS and OS
Santos et al[16]	2014	TRG 3 and 4	TRG 0 and 1 and 2	1.8	3.5	NC	11.1	16.4	NC	78.4	68.1	No	77.4	69.1	No				Good responders have better, non-statistically significant outcomes for DFS and OS
Hav et al[15]	2015	TRG 3 and 4	TRG 0 and 1 and 2							Better in GR		No							Good responders have better, non-statistically significant outcomes for DFS

Where data is not given the overall result is stated. LR: Local recurrence; DR: Distant recurrence; GR: Good responders; PR: Poor responders; NC: No statistical comparison made.

Table 8 Summary of magnetic resonance imaging regression scale available in the literature

mrTRG scale	mrTRG
	(Low no. - More regression)[47]
1	Radiological complete response: no evidence of ever treated tumour
2	Good response (dense fibrosis; no obvious residual tumour, signifying minimal residual disease or no tumour)
3	Moderate response (50% fibrosis or mucin, and visible intermediate signal)
4	Slight response (little areas of fibrosis or mucin but mostly tumour)
5	No response (intermediate signal intensity, same appearances as original tumour)

Table 9 Characteristics of studies reporting on poor response based upon magnetic resonance imaging

Ref.	Year	Chemotherapy protocol	Radiotherapy protocol (Gy)	Surgical procedures	TME	Time to surgery (wk)	Cancer stage Pre neo-adjuvant therapy	Adjuvant therapy
Shihab et al[52]	2011			APR/AR	Y
Patel et al[7] and Siddiqui et al[8]	2011 and 2012			APR/AR	Y
Patel et al[6]	2012			APR/AR	Y		T1/2/3/4, N0/1/2	Y
Yu[53]	2014 (unpublished data from our centre)	Capecitabine, oxaliplatin ± cetuximab	50.4-54		Y		T2/3/4	Y
Yu[53]	2014 (unpublished data from our centre)	Capecitabine, oxaliplatin ± cetuximab	50.4-54		Y		T2/3/4	Y

Table 10 Study definitions of poor response according to magnetic resonance imaging tumour regression grade scales

Ref.	Year	TRG scale used (histological stage based upon)	Scales accurate?	Poor response definition	Total (n)	Poor responders (n)	Average F/up in months	LR (%) 5 yr	DR (%) 5 yr	DFS (%)	OS (%)
Shihab et al[52]	2011	MRI TRG (based on Mandard)	Yes	TRG 4,5	37	17		4	9
Patel et al[5,7]	2012	MRI TRG (based on Dworak)	Yes	TRG 4,5	69	22
Patel et al[6] and Patel et al[7]	2011 and 2012	MRI TRG (based on Dworak)	Yes	TRG 4,5	66	34	60	29		Yr. 5: 31	Yr. 5: 27
Yu[53]	2014 (unpublished data from our centre)	MRI TRG (based on Mandard and Dworak)	Yes	TRG 4,5	210	85				Yr. 3: 52%	Yr. 3: 74%
Yu[53]	2014 (unpublished data from our centre)	MRI TRG (based on Mandard and Dworak)	Yes	TRG 4,5	152	47				Yr. 5: 59%	Yr. 5: 68%

LR: Local recurrence; DR: Distant recurrence; DFS: Disease-free survival; OS: Overall survival; TRG: Tumour regression grade.

Table 11 Study definitions of good response according to magnetic resonance imaging tumour regression grade scales

Ref.	Year	TRG scale used (histological stage based upon)	Scales accurate?	Good response definition	Total (n)	Good responders (n)	Average F/up in months	LR (%) 5 yr	DR (%) 5 yr	DFS (%)	OS (%)
Shihab et al[52]	2011	MRI TRG (based on Mandard)	Yes	TRG 1,2,3	37	20		1	3
Patel et al[6]	2012	MRI TRG (based on Dworak)	Yes	TRG 1,2,3	69	47
Patel et al[5] and Patel et al[7]	2011 and 2012	MRI TRG (based on Dworak)	Yes	TRG 1,2,3	66	32	60	14		Yr. 5: 64	Yr. 5: 72
Yu[53]	2014 (unpublished data from our centre)	MRI TRG (based on Mandard and Dworak)	Yes	TRG 1,2	152	61				DFS, Yr. 5: 83%	DFS, Yr. 5: 90%

LR: Local recurrence; DR: Distant recurrence; DFS: Disease- free survival; OS: Overall survival; TRG: Tumour regression grade.

Table 12 Comparison of outcomes between good and poor responders

Ref.	Year	Local recurrence (LR)	P < 0.05	Distant recurrence (DR)	P < 0.05	Progression disease-free survival (DFS)	P < 0.05	Disease-free survival (DFS)	P < 0.05	Overall survival (OS)	P < 0.05	Conclusion
Shihab et al[52]	2011	Better in GR	No	Better in GR	Yes							Good responders have better, statistically significant outcomes for DR but have better, non significant LR
Patel et al[5] and Patel et al[7]	2011 and 2012	Better in GR	No					Better in GR	Yes	Better in GR	Yes	Good responders have better, statistically significant outcomes for DFS and OS but have better, non significant outcomes for LR
Yu[53]	2014					Better in GR	Yes			Better in GR	Yes	Good responders have better, statistically significant outcomes for DFS and OS

GR: Good responders; NC: No statistical comparison made; DNI: Data not interpretable.