Table 3.

Randomized Controlled Trials: Celiac Plexus Block in Pancreatic Cancer.

Study	Patients	Interventions	Results	Comments	AHRQ Quality Score [36]
Mercadante 1993 [30]	20 pancreatic cancer patients randomized after receiving 1 week of analgesics; 10 received NSAIDS/narcotics, 10 had NCPB via posterior approach	Pts either received analgesic meds and were increased toward goal dosages to obtain VAS <4 ; OR underwent NCPB with 25 ml 75% alcohol bilaterally via posterior percutaneous approach then received analgesics like first group.	Though both groups had significant reduction in VAS, no difference between the two. Significant reduction of opioid use in NCPB group	Randomization method not described, unblinded, small study population Opioid intake reduced for up to 7 weeks post NCPB or until time of death	4.5/10
Lillemoe 1993 [26]	Pts with histologically proven unresectable pancreatic cancer, 72 had placebo, 65 had splanchnicectomy	Intraoperative chemical splanchnicectomy with 20ml 50% alcohol Control: 20 ml NS injection as a placebo	NCPB group: Patients without preop pain had significantly reduced VAS scores and delayed onset of or no subsequent pain Pts with preop pain had both reduced pain and increased survival time.	Randomization method not described, double-blinded Nearly 2/3 of pts with preop pain and relief by NCPB had return of mod – severe pain before death. Data implies 3–4 mo of min to mild pain before return of severe Sx.	8/10
Staats 2001 [33]			Increased longevity in NCPB group; significant negative correlation between postop pain and longevity.		8/10
Kawamata 1996 [28]	21 pancreatic cancer pts in palliative care	NCPB: in 10 pts, 8 ml LA + 15–20 ml 80% alcohol Control: in 11 pts, NSAID-morphine. Increase in dose when VAS ≥ 3/10 SQ morphine equivalent given when unable to take orally	VAS scores significantly lower in NCPB group for first 4 weeks, morphine consumption significantly lower in weeks 4–7. Though QOL scores did not differ significantly, they deteriorated only slightly in CBP group	Randomization method not described, unblinded, small study population	5/10
Polati 1998 [29]	Pts with histologically proven unresectable pancreatic cancer, 12 pts underwent NCPB, 12 pts had pharmacotherapy	NCPB: 6–8 ml of LA + 7 ml of absolute alcohol, Control: 6–8 ml of LA + WHO guidelines of pharmacotherapy	Immediate significant pain relief (in first 48 hours) in NCPB group; but long-term results did not differ between two groups. Reduced opioid need in NCPB group at ¼ and ½ survival time (not significant at ¾ survival time) and thus reduced opioid side effects.	Randomization method not described, double-blinded, small study population	6.5/10
Wong 2004 [27]	Pts receiving noncurative pancreatic surgery were eligible with a NRS of ≥ 3/10. 50 pts/group used in analysis	NCPB: 10 ml LA + 10 ml absolute alcohol Control: sham procedure by SQ and IM LA+ pharmacotherapy Rescue blocks if NRS ≥ 6/10 or intolerable opioid adverse effects	Greater reduction in pain scores in NCPB group but no significant difference in opioid consumption, QOL and survival.	Randomization by calling a central telephone number in blocks of 4 pts per group, double-blinded, small study population. Though a greater number of patients survived in NCPB group, results were not significant.	8/10
Zhang 2008 [37]	56 pts with unresectable pancreatic cancer, 29 pts had CT-guided NCPB, 27 treated with pharmacotherapy	NCPB: 5 ml LA + 20 ml absolute alcohol Control: MS contin	At day 1, 7 and 14 VAS lower in NCPB than control; opioid consumption lower in NCPB group. Though both groups improved, QOL not different between two groups	Randomization method not mentioned, unblinded.	6/10
Johnson 2009 [31]	65 pts (57 pancreatic cancer, 3 gallbladder cancer, 1 bile duct cancer, 1 duodenal cancer, 3 unknown); 18 withdrew or died in 2 months	MM: protocol for opioids CPB: “usually alcohol”, done by various operators TS: done by various operators	No difference in pain relief or opioid consumption between the 3 groups.	Multicenter study, Randomization by telephone in blocks of 3 and stratified by treatment center, tumor type, and current opioid status. Unblinded, Small study population. No standardized injectate for CPB described	7/10

Pts: patients, NS: normal saline, LA: local anesthetic, MS: morphine sulfate, WHO: World Health Organization, Sx: symptoms, NRS: numeric rating scale, SQ: subcutaneous, IM: intramuscular, QOL: quality of life, VAS: visual analogue scale, MM: medical management, TS: thoracic splanchnicectomy.