Using fluorodeoxy-D-glucose-positron emission tomography to monitor neoadjuvant chemotherapy response in sarcoma: A meta-analysis

Table 1 Characteristics of the included studies

	Year	Tumor pathology	Patient No. (% of males)	Age, mean (range), yr	Therapy, regimen	No. (%) of responders	Pathologic criteria	Adopted PET criteria	Timing of PET	Quality scoreb
Im	2012	OS (resectable high-grade)	14 (71)	15 (10-25)	Neo-chemo: various combinations of D, C, M, I and E	5 (36)	Necrosis fraction ≥ 90% (by Salzer-Kuntschik et al)	Post-chemo SUVmax < 3	Before, between the first and second cycles	12
Kim	2011	OS or ESFT	23 (69)	10 (3-19)	Neo-chemo: various combinations, I contained	15 (65)	Viable tumor ≤ 10%	SUV reduction rate ≥ 50% and SUV2 ≤ 2.5	Before, after completion of chemo	10
Bajpai	2011	OS	31 (81)	17 (5-66)	Neo-chemo: C + D	10 (32)	Necrosis fraction ≥ 90%	V1 ≤ 300 mL and SUV2:SUV1 ≤ 0.48	Before, after 3 cycle	13
Dimitrako	2010	STS (non-metastasized)	24 (unclear)	Unclear	Neo-chemo: E + D	14 (58)	Necrosis fraction ≥ 90%	SUVmean+ Influx, value unclear	Before, after 2 cycles	9
Cheon	2009	OS (high-grade)	70 (68)	14a (5-59)	Neo-chemo: M, A and C	33 (47)	Necrosis fraction ≥ 90%	SUV2 < 2 or MVCR > 65% when SUV2 is between 2 to 5	Before, after completion of chemo	12
Hamada	2009	OS	9	51a (20-80)	Neo-chemo: D, C, I and M	5 (56)	Necrosis fraction ≥ 90%	SUV2 < 2.5	Before, after completion of chemo	10
Benz	2009	STS (resectable high-grade)	50 (52)	51a (20-80)	Neo-chemo: mostly I contained	8 (16)	Necrosis fraction ≥ 90%	SUV reduction rate ≥ 35% after the first cycle	Before, after the first cycle, after completion of chemo	12
Benz	2008	STS (high-grade)	20 (50)	49 (19-86)	Neo-chemo: I + D or G + D; some had radiotherapy	6 (30)	Necrosis fraction ≥ 95%	SUVmean changes, value unclear	Before, after completion of chemo	12
Evilevitch	2008	STS (high-grade)	42 (57)	54 (20-86)	Neo-chemo: I in 84%, G in 16% of patients, some had radiotherapy	8 (19)	Necrosis fraction ≥ 95%	SUV reduction rate ≥ 60%	Before, after completion of chemo	13
Iagaru	2008	OS and STS	14 (57)	36 (18-56)	Neo-Chemo: I based	6 (43)	Necrosis fraction ≥ 90%	SUV reduction rate ≥ 27%	Before, after completion of chemo	11
Ye	2008	OS	15 (60)	17a (7-33)	Neo-chemo: M, C, and A	8 (53)	Necrosis fraction ≥ 90%	TBR2/TBR1 < 0.46	Before, after completion of chemo	13
Hawkins	2005	ESFT	34 (unclear)	19 (6-46)	Neo-chemo: V, D, Cy, I and E; some had radiotherapy	25 (74 )	Necrosis fraction ≥ 90%	SUV2 < 2.5	Before, after completion of chemo	10
Hawkins	2002	OS and ES	31 (74)	13 (6-19)	Neo-chemo: D, C, M, with or without I	18 (58)	Necrosis fraction ≥ 90%	SUV2 < 2	Before, after completion of chemo	10
Franzius	2000	OS and ES	17 (76)	13a (5-36)	Neo-chemo: D, M, C and I; some had radiotherapy	15 (88)	Necrosis fraction ≥ 90%	TBR reduction rate > 30%	Before, after completion of chemo	12
Schulte	1999	OS	27 (63)	17a (5-36)	Neo-chemo: D, M, I and C	17 (63)	Necrosis fraction ≥ 90%	TBR2/TBR1< 0.6	Before, after completion of chemo	12

^a: Median age;

^b: Evaluated by QUADAS tool, presented as the number of “Yes” to the fourteen questions of QUADAS criteria. A: Adriamycin; C: Cisplatin, Cy: Cyclophosphamide; D: Doxorubicin, E: Etoposide, G: Gemcitabine; M: Methotrexate, I: Ifosfamide; V: Vincristine; OS: Osteosarcoma; STS: Soft-tissue sarcomas; ES: Ewing sarcoma; ESFT: Ewing sarcoma family of tumors; Neo-chemo: Neoadjuvant chemotherapy; SUV: Standardized uptake value; SUV1: Pre-chemo SUV; SUV2: Post-chemo SUV; MVCR: Metabolic volume change ratio; TBR: Tumor to background ratio; TBR1: Pre-chemo TBR; TBR 2: post-chemo TBR.

Table 2 Absolute numbers from 2 × 2 contingency tables for the included studies

	TP	FN	FP	TN
Im	5	0	1	8
Kim	14	1	1	7
Bajpai1	4	6	0	20
Dimitrako	11	3	1	9
Cheon	31	2	1	36
Hamada	5	0	0	4
Benz-2009	8	0	14	28
Benz-2008	6	0	0	14
Evilevitch	8	0	10	24
Iagaru	3	3	3	5
Ye	8	0	0	7
Hawkins-2005	19	6	5	4
Hawkins-2002	14	4	1	12
Franzius	15	0	0	2
Schulte	17	0	2	8

¹One patient was not taken into calculation due to the unavailabilty of the relevant PET data. TP: True positive; FP: False positive; FN: False negative; TN: True negative.

Table 3 Subgroup analysis based on tumor pathology, patient age groups and chemotherapy regimen

Subgroupcriteria	Characteristics	No. of studies	No. of patients	Se %	P of Se	Sp %	P of Sp	DOR
Tumor pathologya	OS or ES (1-3, 5, 6, 11-15)	10	270	87 (95%CI: 81-92)	> 0.05	91 (95%CI: 84-95)	< 0.05	57.8
	STS (4, 7, 8, 9)	4	55	92 (95%CI: 78-98)		75 (95%CI: 65-83)		47.9
Patient age groupsb	Pediatric patients (1, 2, 13)	3	68	87 (95%CI: 72-96)	> 0.05	90 (95%CI: 73-98)	< 0.01	59.2
	Adult patients (6, 7, 8, 9, 10)	5	135	91 (95%CI: 76-98)		74 (95%CI: 64-82)		25.6
Chemotherapy regimenc	With ifosfamide (2, 6, 7, 9, 10, 12, 14, 15)	8	216	90 (95%CI: 82-95)	> 0.05	70 (95%CI: 61-78)	< 0.01	20.8
	Without ifosfamide (3, 4, 5, 11)	4	139	83 (95%CI: 72-91)		97 (95%CI: 91-100)		103.8

^a[22] not included due to covering both tumor pathology;

^b7 studies not included due to age ranges covering both pediatric and adult patients;

^c3 studies not included [3,20,25] for unclear portion of ifosfamide use. Se: Sensitivity; Sp: Specificity; OS: Osteosarcoma; ES: Ewing sarcoma; DOR: Diagnostic odds ratio.