- Email: [email protected]
Intraarterial adriamycin and lipiodol for inoperable hepatocellular carcinoma: a comparison with intravenous adriamycin
349
Intraarterial adriamycin and lipiodol for inoperable hepatocellular carcinoma: a comparison with intravenous adriamycin
The technique of ‘targeting’ cytotoxic drugs by mixing them v&b the contrast medium lipiodal is now u+dely used in Japan and the Fat East where it has been reported to enhance respa~~e rates in patients with hepatocelbdar carcinoma. In the present study 19 patiettts with this tummtr were treated with iatra-@epatic)artetial adriamycb? (60 mp/m*), at least one course ofwbich was combined with lipiodol (IO-20 ml). Two patients (11%) bad a mmissiott as indicated by a sign&ant fall in serum alpbafetopmtem and there war a reduction of tumcw size in one of &se. The median mwivaf perid was 3 momhs (range l-18)withtbctworespandiogpatients~nggand 12months. Tbisresponseratewasnbettertbantbe figure of 14% seen in 31 eolgccutive patieots treated with intravexms adriamycin at the same dose, and the survival awes ofthctwogoupsofpatien~tserenotsign~tly different. Lipiodol in Combination~~adriamyeiniswtNperiortointravenous adtiamycin administered alone.
Adriamycin is one of the few single agents to exhibit mty degree cd activity against hcpatozellular car&ma (HCC) with a reported mspatse rate of between 9-44% when &ministered systemically (l-6). Ablmugb cambinatiottcbemotltera~utkregbnens, witborwithcmt adtiamycin, have met titb little SumB (7.8). several reports have wggested that administratkm of the drug direztly In
Fii-four patienti with bwperabk HCC, seen on the Liver Unit between January 1987 and September 1989, were treated with adriamycin within 1 month of diagtiais. AU ftdfilled standard criteria for the diagaoJir of Hcc: either a serum alpb&opmtein (AFP) level of greater than500n~~~alinPpnt~ta~ve~~~~~~
Japan and the Far Earf awJministmtion of the radiobgical calbaa medium l&icdol whicit iocalises preferentially to tumottt the, is nmv a anll_es’aM procedure (9,lO). same Ivurkem have aIS0 wad tbc water soule cotmast utxdiuta diatriwate SQdbml meghlmb2e (LJmgrafin) wbicb is dabned to increase the stab@ of the adriamycbulipiodolemtdsion (11,X?). Overtbe lart 3yeats we have tndetin detailed pbarttmmkbtetic studies of the factors bttluenciog tbe disposidon of adriamycin and, ia pa&&r, the effect of cadmiUistntioD of li+dol. we now report on the clinical rcsponsesofthep&etttsidinthesestudies.
hepatic mass demcmsuafcd on either tdttasottnd or computerized tomography (Cl’) aad& diagn&ic histok~gy. Ihe stage of the dbease was clasSad accmdiig to Okuda et al. (13). Brietly, this invdvg a scming system based on ssrum biiimbin comzeturatiot~, albumin mnc.zmratimtandtbeptw.zttaora~ofascit~attdtumoar size.stage1bastbzbcstpmBooris aad stage III the worst. Hepatitis B surface antigen (HBsAg) was detected by aliobnmunoarsay (Austin II, Abbott Laboratmies, lL)_ The tint 31 patients were treated with intravenous and dte next 23 with b&ate&l adriamycin. AU received adriamycin @I m%&) as a bobts iajextion at 3 to 4 week
into fbe hcpltie
artery might be more effective (S-10).
350 intervals. Tbe dose was reduced in a single patient (in the intraanerial group) to 25 m&, tgsauw of an extremely high serum bilimbin concentration (IMl~moUl) and tense ascites. All other patients had B serum bilirubin conmntntion of below 60 ~molll (reference range less than 20 ~mobl), and recaived the full calculated dose of drug. The treatment protocol for those receiving the drug by the imxaanerial route (as ioitially derieed for pharmacakinetic studies) was that each patient would receive two courses of therapy; either adriamydn alone (adm), and then 3 weeks later, adriamycin mixed with lipiodol (adml lip), or vice versa. Four patients subsequently received a third ~wse comprising ad&my&, lipicdol and umgra fir! (adm/lip/uro). This report concerns those 19 patient: who received adriamycin and lipicdol, and excludes four patients who died of progressive disease shortly after their ftrst, ‘adriamycin alone’ treatment, i.e., before receiving the amme combined with lipiodol. A full blocif count, including haemogtobin concentmtion, total white cell cmmt and platelet cmmt, was taken immediately prior to adriamycin administration, and sub. sequently. wherever possible, on dayrl, 14 and 21.
One hundred milligrams of adriamycin was dissolved in 25 ml of normal saline (or urografin, 76%) and mixed thoroughly. When combined with lipiodol the requisite dose (60 mg/m’) was taken up in B sylinge and connected via a tube to a second syringe containing 10 ml (or, when urogmfin was used, 20 ml) of lipiodol. Immediately before administration to the patient, the two agents were extensively mixed by pumping to-and-fro between the syringes at least twenty times. Tbe hepatic artery was selectively catheter&d using the Seldinger technique and conventional rapid sequence films obtafned by injecting nonionic contrast medium. The lipiodolladriamycin mixture was then administered directly into the hepatic artery as a bolus over2 min. ConArmation that the lipiodol bad local&d to the turnour areas was obtained by computed tomography at 1 week (Fig. 1). A response was mnsidered to have warred when there ~88 at least a 30% reduction in hepatamegaly as assessed clinically by distance below the costal margin; at least a 50% reduction is turnour diameter cm serial ultrasound examination in patients with a solitary mass; or at
351 least a 50% fall in the serum AFP level (where the pretreatment level was greater than 250 @ml) provided that there was no contradictory evidence from the other two criteria. Stable disease was defined as a less than 50% decrease in she of the tumour as measured by Cf or less than 50% decrease in serum APP over 4 weeks. Increase in AFP levels, liver size or development of new metastasis was considered to indicate progressive disease. Response was measured from the date of the initial treatment until the time of diseirse progression. Survival time was calculated from the time of diagnosis and survival cwves were calculated using the Kaplan-Meier method (14).
stage of the disease at presentation (Table I). AU tattour% were greater than 3 cm in diameter. Among those receiving intraanerial cbemotberapy, ten patients received one course (admtlip in six; admllipiuro in four). Five received two causes (adm followed by w&/lip), and the remaining four patients had three courses (adm alone, adtip and
died within I week of starting treatment, one underwent liver transplantation at 3 weeks and seven were loat tea follow-up after returning to their own country. These patients are includedinthesurvival analysisupuntil the time they were last knovm to be alive. Two of the 18 (11%) evaluable patients in the intraarterial group undetwent a response, surviving for 8 and 12 months. In the first patient who eventually died after a massive variceal baemonbag~ there was a reduction in tumoursizeof greaterthan50% @measured byCX) and a fall in APP levels from 2465 to 629 @ml aher three cowsesofhearmem.TheseeondpatiemhadafaltinAm from 2310 to 11Ml nglml aher the second course. This patient is stil! zXv= and asymptomatic at 12 months folkkq, diagnosis. although with a progressively rising AFP. These two patients a&n showed improvement in terms of relief of their presenting symptoms; right upper quadrant pain in the Fust and severe ptitus in the second. Among those with no objective evidence of response, tw had symptomatic improvement, with disappearance of the fever in one and of right upper qua&ant pain in the other. Three patients had stable disease with no increase in the turnour size or AFP level at 6.9 and 11 months. The remaining 14 pattents showed turnour progressioti as dczumented radiologically and/or by serial APP estimations. Among thw in the inttavenous gmup, 14 patfen& bad
admiXplum). Of the 19 patients 18 were evaluable for resprmse and one c&d not be further traced. Twenty-two of the 31 patients who had intravenous adriamycin Matment were evaluable, and nine were unevaluable. One
one course of treatment, seven had two courses, five had three courses, two had six caxses, one had seven uluises and two had nine cauees. Three patients (14%) reqtowSed, wrviving 3.13 and 14 manths after the diagmnk. The
Res*ts The two patient groups (intraarterial and intravenous) were well matched with respect to all features relating to prognosis including age, serum bilirubin concentration, presence or absence of cirrhosis, performance stws and
aped severe, and one mild mucositis in the intravenous treatment group. Three patients in the intravenous group and five in the intraarterial group (three, with, two without mixing with liiiodol) had fever and rigon, within 6 h of drug administration, whiih lasted less than 1 h and resolved without any treatment. Seven patients treated with intravenous sdriamycin and four with intraarterial, cornplained of nausea following drug administration and this was associated with vomiting in four and two, respectiveIY. No complicatio” wet. observed due to trzmsfemoral hepatic artery catheterization. There were no instances of adriamyci” related cardiac toxicity.
first patient had a reduction in the turnout siie of greater than SO%, associated with a fall in AFP from 912 to 690 “g/ml within 3 weeks but died suddenly, 3 months after starting treatment, following a variceal haemorrhage. The other two patients had a reduction in AFF’ from 106 OKl to 30 4w “g/ml and from 252 to 113 nglml. Apart from one of the responders who reported an increase in well-being no other patient in this group showed any
This study represents the largest series of patients so fax treated in the West, and whilst most patients received one course of intraarterial adriantycin alone, all received in addition, at least one course of the combined prepamtiw. Furthermore, this is the iirst trial which permits dimct comparison of the ef&xy of the same regimen pdmioistered intravenously or intraatterially, in matched group
symptomatic improvement. Five patients had stable disease and three of these are alive and well at 6,7 and 12 months. The remaining 14 patients had progressive dik ease with no evidence of response to treatment. There was no significant difference in the survival curves of the two groups (Fig. 2) and the median survival was almost identical (intraarterial group, 3 months (range, l-18); intravenous group, 4 months (rrmge, 1-19)). The survival times of the responders, in both groups were greater than in the non-responders (intraarterial group: responders, 8 and 12 months, “on-responders, media” 2 months (range, I-18); intravenous group: responders 3,13 and 14 months, “on-responders, median
of patients. We meld detect “o advantage for intraarterial adria. mytin with lipiodol over intravenous adriamyrin either in terms of response rate m survival, a”d although the study was not randomised, the two groups were well matched in relation to all pmg”ostic variables. These results are, per-
4 months (range, 1-19)). Regardless of the method of administration, those without cirrhosis had a Longer media” survival (7 months, range 2-18) the” those with cirrhosis (4 months, range 1-19). The response rate was also higher in non-cirrhotics (two cut of ten (20%) compared to three out of 37 (7.69%)).
use of lipiodol and adriamycin itt patients with HCC and concluded that it WBS safe and effective (11.15-19). Whilst the precise details of administration and fomntlatimt have differed, and the number of OOUM ad&stered is often not reported. most have followed this form of treatment with emboliition of the feeding artery uping Relfoam fl?,lS-19). This leads to difficultv io delineatinP the effect of the individual forms of treatment. Oux data suggests that where such therapy is effective in caw tnmour necrosis, it is probably the embolisstion rather the” the chemotherapy which is responsible. This is pattialarly likely in view of the fact that the dose of adripmyein administered in the present study was at least twice that m ported by other workers. There groups have also claimed that this form of thcra-
Toxici~ Whilst there was marked neutropaenia (maximal on day 14) and thrombocytopaenia (maximal on day 7) in all patients, there was no significant difference between the two groups and only one episode of septicaemia which responded rapidly to appropriate antibiotic therapy (14). All patients developed some degree of alopecia regardless of the method of administration. One patient devel-
haps, not surprising in view of our pharmacokinctic smdies w&h suggested that the coadministtatian of llpiado with adriamyci” had no observable effeei on peripheral drug levels. terminal half-life or toxicity, when compared to treatment with either intravenous or intraarterial adrlamycin alone. Several groups from the Fat East have d&bed
the
353 py probably leads to an increase in suvival (11,12,17). However, without a prospective controlled trial this is difficult to prove, particularly in view of the prolonged SWviva1 reported in patients with small turnours. Okuda et al. (20) reported survival of 90% at 1 year even in the atsence of any f%?abnenf, and it is noteworthy that in the experience of Takayasu et al. (17). on multiple regression analysis. the size of the turnour rather than any method of administration (i.e., with or without lipiodol) was the only factor influencing survival. Some authors have considered
that the use of lip&&l perse causes therapeutic embolisadon of the furnow (10.21) though this has been disputed (22).
Intraarterial adriamycin and lipiodol for inoperable hepatocellular carcinoma: a comparison with intravenous adriamycin
The technique of ‘targeting’ cytotoxic drugs by mixing them v&b the contrast medium lipiodal is now u+dely used in Japan and the Fat East where it has been reported to enhance respa~~e rates in patients with hepatocelbdar carcinoma. In the present study 19 patiettts with this tummtr were treated with iatra-@epatic)artetial adriamycb? (60 mp/m*), at least one course ofwbich was combined with lipiodol (IO-20 ml). Two patients (11%) bad a mmissiott as indicated by a sign&ant fall in serum alpbafetopmtem and there war a reduction of tumcw size in one of &se. The median mwivaf perid was 3 momhs (range l-18)withtbctworespandiogpatients~nggand 12months. Tbisresponseratewasnbettertbantbe figure of 14% seen in 31 eolgccutive patieots treated with intravexms adriamycin at the same dose, and the survival awes ofthctwogoupsofpatien~tserenotsign~tly different. Lipiodol in Combination~~adriamyeiniswtNperiortointravenous adtiamycin administered alone.
Adriamycin is one of the few single agents to exhibit mty degree cd activity against hcpatozellular car&ma (HCC) with a reported mspatse rate of between 9-44% when &ministered systemically (l-6). Ablmugb cambinatiottcbemotltera~utkregbnens, witborwithcmt adtiamycin, have met titb little SumB (7.8). several reports have wggested that administratkm of the drug direztly In
Fii-four patienti with bwperabk HCC, seen on the Liver Unit between January 1987 and September 1989, were treated with adriamycin within 1 month of diagtiais. AU ftdfilled standard criteria for the diagaoJir of Hcc: either a serum alpb&opmtein (AFP) level of greater than500n~~~alinPpnt~ta~ve~~~~~~
Japan and the Far Earf awJministmtion of the radiobgical calbaa medium l&icdol whicit iocalises preferentially to tumottt the, is nmv a anll_es’aM procedure (9,lO). same Ivurkem have aIS0 wad tbc water soule cotmast utxdiuta diatriwate SQdbml meghlmb2e (LJmgrafin) wbicb is dabned to increase the stab@ of the adriamycbulipiodolemtdsion (11,X?). Overtbe lart 3yeats we have tndetin detailed pbarttmmkbtetic studies of the factors bttluenciog tbe disposidon of adriamycin and, ia pa&&r, the effect of cadmiUistntioD of li+dol. we now report on the clinical rcsponsesofthep&etttsidinthesestudies.
hepatic mass demcmsuafcd on either tdttasottnd or computerized tomography (Cl’) aad& diagn&ic histok~gy. Ihe stage of the dbease was clasSad accmdiig to Okuda et al. (13). Brietly, this invdvg a scming system based on ssrum biiimbin comzeturatiot~, albumin mnc.zmratimtandtbeptw.zttaora~ofascit~attdtumoar size.stage1bastbzbcstpmBooris aad stage III the worst. Hepatitis B surface antigen (HBsAg) was detected by aliobnmunoarsay (Austin II, Abbott Laboratmies, lL)_ The tint 31 patients were treated with intravenous and dte next 23 with b&ate&l adriamycin. AU received adriamycin @I m%&) as a bobts iajextion at 3 to 4 week
into fbe hcpltie
artery might be more effective (S-10).
350 intervals. Tbe dose was reduced in a single patient (in the intraanerial group) to 25 m&, tgsauw of an extremely high serum bilimbin concentration (IMl~moUl) and tense ascites. All other patients had B serum bilirubin conmntntion of below 60 ~molll (reference range less than 20 ~mobl), and recaived the full calculated dose of drug. The treatment protocol for those receiving the drug by the imxaanerial route (as ioitially derieed for pharmacakinetic studies) was that each patient would receive two courses of therapy; either adriamydn alone (adm), and then 3 weeks later, adriamycin mixed with lipiodol (adml lip), or vice versa. Four patients subsequently received a third ~wse comprising ad&my&, lipicdol and umgra fir! (adm/lip/uro). This report concerns those 19 patient: who received adriamycin and lipicdol, and excludes four patients who died of progressive disease shortly after their ftrst, ‘adriamycin alone’ treatment, i.e., before receiving the amme combined with lipiodol. A full blocif count, including haemogtobin concentmtion, total white cell cmmt and platelet cmmt, was taken immediately prior to adriamycin administration, and sub. sequently. wherever possible, on dayrl, 14 and 21.
One hundred milligrams of adriamycin was dissolved in 25 ml of normal saline (or urografin, 76%) and mixed thoroughly. When combined with lipiodol the requisite dose (60 mg/m’) was taken up in B sylinge and connected via a tube to a second syringe containing 10 ml (or, when urogmfin was used, 20 ml) of lipiodol. Immediately before administration to the patient, the two agents were extensively mixed by pumping to-and-fro between the syringes at least twenty times. Tbe hepatic artery was selectively catheter&d using the Seldinger technique and conventional rapid sequence films obtafned by injecting nonionic contrast medium. The lipiodolladriamycin mixture was then administered directly into the hepatic artery as a bolus over2 min. ConArmation that the lipiodol bad local&d to the turnour areas was obtained by computed tomography at 1 week (Fig. 1). A response was mnsidered to have warred when there ~88 at least a 30% reduction in hepatamegaly as assessed clinically by distance below the costal margin; at least a 50% reduction is turnour diameter cm serial ultrasound examination in patients with a solitary mass; or at
351 least a 50% fall in the serum AFP level (where the pretreatment level was greater than 250 @ml) provided that there was no contradictory evidence from the other two criteria. Stable disease was defined as a less than 50% decrease in she of the tumour as measured by Cf or less than 50% decrease in serum APP over 4 weeks. Increase in AFP levels, liver size or development of new metastasis was considered to indicate progressive disease. Response was measured from the date of the initial treatment until the time of diseirse progression. Survival time was calculated from the time of diagnosis and survival cwves were calculated using the Kaplan-Meier method (14).
stage of the disease at presentation (Table I). AU tattour% were greater than 3 cm in diameter. Among those receiving intraanerial cbemotberapy, ten patients received one course (admtlip in six; admllipiuro in four). Five received two causes (adm followed by w&/lip), and the remaining four patients had three courses (adm alone, adtip and
died within I week of starting treatment, one underwent liver transplantation at 3 weeks and seven were loat tea follow-up after returning to their own country. These patients are includedinthesurvival analysisupuntil the time they were last knovm to be alive. Two of the 18 (11%) evaluable patients in the intraarterial group undetwent a response, surviving for 8 and 12 months. In the first patient who eventually died after a massive variceal baemonbag~ there was a reduction in tumoursizeof greaterthan50% @measured byCX) and a fall in APP levels from 2465 to 629 @ml aher three cowsesofhearmem.TheseeondpatiemhadafaltinAm from 2310 to 11Ml nglml aher the second course. This patient is stil! zXv= and asymptomatic at 12 months folkkq, diagnosis. although with a progressively rising AFP. These two patients a&n showed improvement in terms of relief of their presenting symptoms; right upper quadrant pain in the Fust and severe ptitus in the second. Among those with no objective evidence of response, tw had symptomatic improvement, with disappearance of the fever in one and of right upper qua&ant pain in the other. Three patients had stable disease with no increase in the turnour size or AFP level at 6.9 and 11 months. The remaining 14 pattents showed turnour progressioti as dczumented radiologically and/or by serial APP estimations. Among thw in the inttavenous gmup, 14 patfen& bad
admiXplum). Of the 19 patients 18 were evaluable for resprmse and one c&d not be further traced. Twenty-two of the 31 patients who had intravenous adriamycin Matment were evaluable, and nine were unevaluable. One
one course of treatment, seven had two courses, five had three courses, two had six caxses, one had seven uluises and two had nine cauees. Three patients (14%) reqtowSed, wrviving 3.13 and 14 manths after the diagmnk. The
Res*ts The two patient groups (intraarterial and intravenous) were well matched with respect to all features relating to prognosis including age, serum bilirubin concentration, presence or absence of cirrhosis, performance stws and
aped severe, and one mild mucositis in the intravenous treatment group. Three patients in the intravenous group and five in the intraarterial group (three, with, two without mixing with liiiodol) had fever and rigon, within 6 h of drug administration, whiih lasted less than 1 h and resolved without any treatment. Seven patients treated with intravenous sdriamycin and four with intraarterial, cornplained of nausea following drug administration and this was associated with vomiting in four and two, respectiveIY. No complicatio” wet. observed due to trzmsfemoral hepatic artery catheterization. There were no instances of adriamyci” related cardiac toxicity.
first patient had a reduction in the turnout siie of greater than SO%, associated with a fall in AFP from 912 to 690 “g/ml within 3 weeks but died suddenly, 3 months after starting treatment, following a variceal haemorrhage. The other two patients had a reduction in AFF’ from 106 OKl to 30 4w “g/ml and from 252 to 113 nglml. Apart from one of the responders who reported an increase in well-being no other patient in this group showed any
This study represents the largest series of patients so fax treated in the West, and whilst most patients received one course of intraarterial adriantycin alone, all received in addition, at least one course of the combined prepamtiw. Furthermore, this is the iirst trial which permits dimct comparison of the ef&xy of the same regimen pdmioistered intravenously or intraatterially, in matched group
symptomatic improvement. Five patients had stable disease and three of these are alive and well at 6,7 and 12 months. The remaining 14 patients had progressive dik ease with no evidence of response to treatment. There was no significant difference in the survival curves of the two groups (Fig. 2) and the median survival was almost identical (intraarterial group, 3 months (range, l-18); intravenous group, 4 months (rrmge, 1-19)). The survival times of the responders, in both groups were greater than in the non-responders (intraarterial group: responders, 8 and 12 months, “on-responders, media” 2 months (range, I-18); intravenous group: responders 3,13 and 14 months, “on-responders, median
of patients. We meld detect “o advantage for intraarterial adria. mytin with lipiodol over intravenous adriamyrin either in terms of response rate m survival, a”d although the study was not randomised, the two groups were well matched in relation to all pmg”ostic variables. These results are, per-
4 months (range, 1-19)). Regardless of the method of administration, those without cirrhosis had a Longer media” survival (7 months, range 2-18) the” those with cirrhosis (4 months, range 1-19). The response rate was also higher in non-cirrhotics (two cut of ten (20%) compared to three out of 37 (7.69%)).
use of lipiodol and adriamycin itt patients with HCC and concluded that it WBS safe and effective (11.15-19). Whilst the precise details of administration and fomntlatimt have differed, and the number of OOUM ad&stered is often not reported. most have followed this form of treatment with emboliition of the feeding artery uping Relfoam fl?,lS-19). This leads to difficultv io delineatinP the effect of the individual forms of treatment. Oux data suggests that where such therapy is effective in caw tnmour necrosis, it is probably the embolisstion rather the” the chemotherapy which is responsible. This is pattialarly likely in view of the fact that the dose of adripmyein administered in the present study was at least twice that m ported by other workers. There groups have also claimed that this form of thcra-
Toxici~ Whilst there was marked neutropaenia (maximal on day 14) and thrombocytopaenia (maximal on day 7) in all patients, there was no significant difference between the two groups and only one episode of septicaemia which responded rapidly to appropriate antibiotic therapy (14). All patients developed some degree of alopecia regardless of the method of administration. One patient devel-
haps, not surprising in view of our pharmacokinctic smdies w&h suggested that the coadministtatian of llpiado with adriamyci” had no observable effeei on peripheral drug levels. terminal half-life or toxicity, when compared to treatment with either intravenous or intraarterial adrlamycin alone. Several groups from the Fat East have d&bed
the
353 py probably leads to an increase in suvival (11,12,17). However, without a prospective controlled trial this is difficult to prove, particularly in view of the prolonged SWviva1 reported in patients with small turnours. Okuda et al. (20) reported survival of 90% at 1 year even in the atsence of any f%?abnenf, and it is noteworthy that in the experience of Takayasu et al. (17). on multiple regression analysis. the size of the turnour rather than any method of administration (i.e., with or without lipiodol) was the only factor influencing survival. Some authors have considered
that the use of lip&&l perse causes therapeutic embolisadon of the furnow (10.21) though this has been disputed (22).