Copyright
©2014 Baishideng Publishing Group Inc.
World J Gastroenterol. Jul 14, 2014; 20(26): 8482-8490
Published online Jul 14, 2014. doi: 10.3748/wjg.v20.i26.8482
Published online Jul 14, 2014. doi: 10.3748/wjg.v20.i26.8482
Table 1 Characteristics of the identified studies
Ref. | Year of publication | Country | Inclusion period | No. of patients | Study design | Follow-up median (95%CI), mo |
Spratlin et al[9] | 2004 | Canada | 1998–2002 | 21 | RS | NR |
Giovannetti et al[10] | 2006 | Italy | 2001–2004 | 1021 | RS | 11.2 (0.4–32.1) |
Farrell et al[12] | 2009 | United States | 1998–2002 | 91 | Post hoc2 | NR |
Maréchal et al[3] | 2009 | Belgium | 2000–2003 | 45 | RS | 21.9 (3.3–107.4) |
Fujita et al[8] | 2010 | Japan | 1992–2007 | 70 | RS | 15.7 (0.5–114) |
Maréchal et al[1] | 2012 | Belgium | 1996–2009 | 234 | RS | 55.7 (46.4–61.2) |
Kawada et al[2] | 2012 | Japan | 2002–2007 | 63 | RS | 31 |
Morinaga et al[21] | 2012 | Japan | 2006–2008 | 27 | RS | NR |
Murata et al[22] | 2012 | Japan | 2005–2010 | 93 | RS | 15 (3.5–57.2) |
Nakagawa et al[23] | 2012 | Japan | 2002–2011 | 109 | RS | 39.7 (2–122) |
Total: 855 |
Table 2 Characteristics of the identified studies
Ref. | Age median, yr (range) | Sex m/tot n (%) | hENT1 method | Chemotherapy | Radiation dose Gy/(Gy/frac) | Outcome measurement | Quality REMARK |
Spratlin et al[9] | 58 (51-64)1 | 11 (52) | IHC | Pall Gem | No | OS | 10 |
Giovannetti et al[10] | 65 (22-83) | 53 (50) | RT-PCR | Pall Gem Adj Gem | 45 | OS, DFS, TTP, RR | 12 |
Farrell et al[12] | 53/63/652 | 45 (49) | IHC | Adj Gem | 50.4 | OS, DFS, Tox | 17 |
Maréchal et al[3] | 56 (34-83) | 23 (51) | IHC | Adj Gem | 40-50.4 | OS, DFS, Tox | 13 |
Fujita et al[8] | 65 (36-86) | 42 (60) | RT-PCR | Adj Gem or Resection only | No | OS, DFS | 12 |
Maréchal et al[1] | NR | 129 (53) | IHC | Adj Gem | 50.4 | OS | 15 |
Kawada et al[2] | - (41-81) | 33 (52) | IHC | Neo Gem Adj 5-FU | 50/2 | DSS | 9 |
Morinaga et al[21] | 64 (45-74) | 17 (63) | IHC | Adj Gem | No | OS, DFS | 12 |
Murata et al[22] | 68 (44-87) | 38 (69) | IHC | Neo Gem Adj Gem | 45/2 | OS, DFS, RR | 13 |
Nakagawa et al[23] | 67 (41-83) | 52 (48) | IHC | Adj Gem + S1 | No | OS, DFS | 13 |
Table 3 hENT1 expression levels, cut-offs and grouping
Ref. | Method | Grading | Reference cells | Groups (n) |
Spratlin et al[9] | IHC | 0-2 based on relative intensities of staining. | Langerhans cells, lymphocytes. | Dichotomized: |
0 = absence of staining | Low = 0 (12) | |||
1 = intermediate staining | High = 1 and 2 (9) | |||
2 = most intense staining | ||||
Giovannetti et al[10] | RT-PCR | Gene-expression ratio with GAPDH, expressed as tertiles | Gene expression tertiles: | |
GAPDH/target gene ratio | Low < 1.06 (27) | |||
Intermediate 1.06-1.38 (28) | ||||
High ≥ 1.38 (26) | ||||
Dichotomized: | ||||
By medians | ||||
Low < 1.23 (44) | ||||
High ≥ 1.23 (37) | ||||
Farrell et al[12] | IHC | Based on relative intensities. | Lymphocytes | Dichotomized: |
High = strong reactivity in > 50% of neoplastic cells. | No (18)1 | |||
No = no staining in > 50% | vs | |||
Low = all cases between High and No. | Low/high (73)1 | |||
Maréchal et al[3] | IHC | 0-3 based on staining intensities | Langerhans cells | Dichotomized: |
0 = no staining | Lymphocytes | Low < 80 (26) | ||
1 = weakly positive | (final score) | |||
2 = moderately positive | ||||
3 = strongly positive | High = ≥ 80 (19) | |||
Final score calculated: multiplying intensity score and the percentage of the specimen. Weighted score 0-300 | ||||
Fujita et al[8] | RT-PCR | Level of mRNA calculated from standard curve constructed with total RNA from Capan-1, a human pancreatic cancer cell line | mRNA split into high/low groups using recursive descent partitioning. Cut-off 0.5 | |
Low (26)1 | ||||
High (14) | ||||
Maréchal et al[1] | IHC | 0-2 based on staining intensities | Lymphocytes | Dichotomized: |
Quantified as Farrell | Low/moderate (136)1 | |||
High (86)1 | ||||
Kawada et al[2] | IHC | 0-2 based on staining intensities. | Langerhans cells | Negative = 0-1 (41) |
1 = same intensity as control. | Positive = 2 (22) | |||
Morinaga et al[21] | IHC | Staining intensity and percentage of positive tumor cells scored and given a hENT1-score by calculating the two | Low = hENT1 score 0-3 (11) | |
Staining 0-3 where | High = hENT1 score 4-6 (16) | |||
0 = no | ||||
1= weakly pos | ||||
2 = moderately pos | ||||
3 = strongly pos | ||||
Percentage: | ||||
0 = no positive | ||||
1 ≤ 50% positive cells | ||||
2 = 50%-80% positive cells | ||||
3 = ≥ 80% | ||||
Murata et al[22] | IHC | Staining intensity + extent of positive staining | Langerhans cells | Dichotomized: |
Intensity: | Negative = low and intermediate (16) | |||
0 = no staining | ||||
1 = weakly positive | Positive = high (39) | |||
2 = moderately positive | ||||
3 = strongly positive | ||||
Extent staining: | ||||
High = score 3 > 50% cells | ||||
Low = score 0 or 1 > 50% | ||||
Intermediate = all others | ||||
Nakagawa et al[23] | IHC | Staining intensities: | Langerhans cells | Low = grade 0 or 1 in > 50% (31) |
0 = not stained | High = grade 2 or 3 in > 50% of cells (78) | |||
1 = faintly stained | ||||
2 = weakly stained | ||||
3 = as strongly as islet cells |
Table 4 Results
Ref. | Median survival all patients (95%CI) | OS | DFS | Main conclusions | ||
Univariate analysis median (95%CI) or HR (95%CI) P-value | Multivariate analysis HR (95%CI) | Univariate analysis median (95%CI) or HR (95%CI) P-value | Multivariate analysis HR (95%CI) | |||
Spratlin et al[9] | 11.01 (6.8-17.5) | (mo): | NR | Pat with detectable hENT1 had sig longer OS compared with pat with low hENT1 | ||
High = 13 (4.2-20.4) | ||||||
5.01 (2.8-12.2) | Low = 4 (1.5-6.9) | |||||
P = 0.01 | ||||||
Giovannetti et al[10] | 13.3 (10.9-15.7) | (mo): | Low = 5.34 (2.28-12.50) | Palliative (mo): | hENT1 expression was significantly correlated with outcome - pat with high hENT1 had longer OS | |
Low = 8.48 (7.01-9.95) | Low = 5.85 (2.75-8.95) | |||||
Inter = 15.74 (13.84-17.63) | Inter = 1.07 (0.46-2.49) | Inter = 10.09 (9.63-10.54) | ||||
High = 25.69 (17.64-33.74) | High = 1 | High = 12.68 (2.89-22.47) | ||||
P ≤ 0.001 | P < 0.0001 | P = 0.02 | ||||
2 groups: | ||||||
2 groups: | HR = 4.21 | Adjuvant (mo): | ||||
Low = 12.42 (8.18-16.66) | P ≤ 0.001 | Low = 9.26 (3.86-14.67) | ||||
High = 22.34 (16.34-28.34) | Inter = 12.91 (9.31-16.51) | |||||
P ≤ 0.001 | High = 20.43 (13.27-27.60) | |||||
P ≤ 0.01 | ||||||
Farrell et al[12] | NR | (HR): | Low/high = 0.40 (0.22-0.75) | (HR): | Low/high = 0.39 (0.21-0.73) | hENT1 expression was ass with longer OS, DFS in pat receiving gem. hENT1 is a relevant predictive marker for gem outcome |
Low/High = 0.51 (0.29-0.91) | Low/High = 0.57 (0.32-1.001) | |||||
No = 1 | No = 1 | No = 1 | No = 1 | |||
P = 0.02 | P = 0.03 | P = 0.05 | P = 0.003 | |||
Maréchal et al[3] | 21.9 (3.3-107.4) | (HR): | High = 1 | (HR): | High = 1 | Pat with high hENT1 had sig longer OS and DFS compared to low hENT1 |
High = 1 | Low = 3.42 (1.44-8.81) | High = 1 | Low = 3.17 (1.43-6.73) | |||
Low = 3.88 (1.78-8.92) | P = 0.0005 | Low = 3.55 (1.65-7.63) | P = 0.0004 | |||
P = 0.0007 | P = 0.02 | |||||
Fujita et al[8] | NR | (mo): | (RR): | (mo): | NR | Low hENT1 ass with shorter OS in gem-group |
High = 45 | Low = 2.980 (0.964-10.86) | High = 25 | ||||
Low = 16.5 | Low = 8 | |||||
P = 0.011 | P = 0.2 (not sig) | P = 0.11 (not sig) | ||||
Maréchal et al[1] | 32.0 (26.4-34.3) | (HR): | n = 2222 | NR | NR | High hENT1 predicts longer OS in pat treated with adj gem. Absence of gem - hENT1 lacks prognostic value |
High = 0.43 (0.29-0.63) | High = 0.34 (0.22-0.53) | |||||
(GEM-group) | Low/Mod = 1 | Low/Mod = 1 | ||||
P < 0.0001 | P < 0.0001 | |||||
Kawada et al[2] | NR | Positive vs negative | Positive/negative | NR | NR | DSS tended to be better in the hENT1-neg group but not statistically sig |
P = 0.352 | P = 0.503 | |||||
Morinaga et al[21] | NR | (mo): | Low = 1 | (mo): | Low = 1 | High hENT1 sig ass with longer OS in pat receiving adj gem after resection |
Low = 11.8 (6.9-16.6) | High = 0.327 (0.128-0.835) | Low = 7.3 (3.6-11.1) | High = 0.558 (0.214-1.452) | |||
High = 22.2 (11.5-32.9) | High = 9.3 (4.2-14.5) | |||||
P = 0.024 | P = 0.019 | P = 0.022 | P = 0.232 | |||
(HR): | (HR): | |||||
Low = 1 | Low = 1 | |||||
High = 0.366 (0.148-0.906) | High = 0.362 (0.146-0.898) | |||||
P = 0.030 | P = 0.028 | |||||
Murata et al[22] | 24.3 | (HR): | Positive = 1 | (HR): | Positive = 1 | Sig longer OS, RFS in pat with pos hENT1 |
Positive = 1 | Negative = 3.15 (1.35-7.37) | Positive = 1 | Negative = 1.76 (0.85-3.66) | |||
Negative = 3.04 (1.45-6.37) | Negative = 2.34 (1.22-4-47) | |||||
P = 0.0037 | P = 0.008 | P = 0.011 | P = 0.129 | |||
Nakagawa et al[23] | OS: 34.9 | (5y-SR %): | High = 1 | (5y-SR %): | High = 1 | hEN1 expression is predictive of the efficacy of adj gem-based chemotherapy after resection |
DFS: 17.8 | High = 38 | Low = 3.16 (1.65-6.06) | High = 30 | Low = 2.70 (1.52-4.83) | ||
Low = 13 | Low = 17 | |||||
P = 0.001 | P = 0.001 | P = 0.004 | P = 0.001 |
- Citation: Nordh S, Ansari D, Andersson R. hENT1 expression is predictive of gemcitabine outcome in pancreatic cancer: A systematic review. World J Gastroenterol 2014; 20(26): 8482-8490
- URL: https://www.wjgnet.com/1007-9327/full/v20/i26/8482.htm
- DOI: https://dx.doi.org/10.3748/wjg.v20.i26.8482