A simplified technic for regional perfusion

A Simplified REPORT

Technic

for Regional

OF ITS USE IN A SERIES OF TWENTY-THREE OF THE HEAD AND NECK

Perfusion

CASES OF MALIGNANCIES

M. GOLOMB, M.D., ALDEN B. HALL, M.D., KENNETH R. Cox, F.R.c.s., F.R.A.c.s., CHARLES F. SCHETLIN, M.D., STEPHEN LA. GUMPORT, M.D. AND JANE C. WRIGHT, M.D., New York, New l’ork

FREDERICK

From tbe Department oj Surgery, New York University Medical Center, Bellevue Hospital, and University Hospital, New York, New York. Tbis study was supported by grants from tbe National Cancer Institute of tbe National Institutes of Health, U. S. Public Healtb Service, Betbesda, Maryland, and by tbe Sbrine Tuberculosis and Cancer Research Foundation.

1957 regional perfusion of human cancers with chemotherapeutic agents was first described [I]. Since then several investigators have reported on their experience with the technic in tumors of the head area [2-41. The methods employed have been essentially those developed by Creech and his co-workers [T]. These incIude the use of a heart-lung machine for maintaining an extracorporeal vascular circuit through the tumor-bearing area. In an attempt to overcome some of the technical problems and perhaps increase the versatility of the perfusion concept a simplified technic has been developed. With it the oxygenator, blood for priming, and the heavy equipment are eliminated but the essential principles of perfusion therapy are maintained.

I

N

Since December 1958, sixty-one perfusions have been performed for neoplasms of various areas. Twenty-eight have been by the simplified, nonoxygenated technic. (Table I.) This report concerns the technic and its use in a series of twenty-three perfusions of the head and neck area. METHOD The standard perfusion technics as described elsewhere [2-s] were employed for the oxygenated perfusions. A sigmamotor pump and disposable plastic, bubble oxygenator (Abbott or Travenol) were used. The nonoxygenated perfusions as described later were performed in a similar manner using a hand-powered pump. Equipment. The nonoxygenated perfusion system used is a single unit containing pump, filter and tubing. (Fig. I.) It is sterile and kept in the operating field, hand operated and disposable. The pump is a plastic bulb equipped with metal bat1 vaIves permitting unidirec-

TABLE I RESULTS OF ALL PERFUSIONS ACCORDING TO USE OR LACK

OF

USE

OF

OXYGEN

II

I

Oxygenated

_

Diagnosis Positive

I

Negative

Adjuvant

Total

Squamous cell carcinoma. Adenocarcinoma Melanoma.............. Sarcoma. GIiobIastoma. TotaI.

15 ; 2 0

23

6 American

Journal

oj Surgery,

Volume

102, December

1961

GoIomb,

HaII, Cox, Schetlin,

FIG. I. Pump system for nonoxygenated regiona perfusion. The apparatus in the Iower part of the iIIustration is the hand-activated pump, fiIter and tubing. The vascuIar catheters or cannulas, not shown, attach to the covered tips at the bottom of the picture. The direction of bIood fIow is counter-clockwise. The calibrated container and attached tubing are a flow meter which can be pIugged into the perfusion circuit at the adapters.

tionaI flow. The afferent and efferent tubing have adapters for injection of drugs and aspiration of bIood sampIes. A metal sieve fiIter is interposed on the afferent Iine. The tubing has adapters for standard needIes, cannuIas and catheters. FIowrates up to 250 cc./ minute can be achieved. A flow meter may be attached to the pump for tIow rate determinations during the perfusion. Technic. General anesthesia is induced and maintained by endotrachea1 technics. A short, diagona1 incision is made aIong the anterior border of the sternocleidomastoid muscIe and the interna jugular vein and externa1 carotid artery are isolated. The artery cannulated depends upon the area to be perfused and may include the external carotid itseIf or any one or combination of its branches. It is not necessary to give heparin intravenousIy. The extracorporea1 circuit is primed with a mixture of 20 mg. heparin and IOO cc. physioIogic saIine soIution. The interna juguIar vein is cannuIated beIow the junction of the common faciaI vein. It is temporarily occIuded above this junction and again beIow the site of cannula840

Gumport

and Wright

tion with umbiIica1 tape tourniquets. A No. 14 or 16 French pIastic catheter or No. 13 or 15 gauge bIunt-tipped, meta cannuIa may be used and the prepared and primed tubing is then attached. The external carotid artery is occluded with a buIIdog or Pott’s clamp proxima1 to the site of cannuIation. The stump of the thyroid artery may be used or the externa1 carotid artery may be incised and Iater repaired or ligated. A meta1, bIunt-tipped cannuIa with a soIdered knob encircling its tip is used. A Iigature pIaced over the vesse1 keeps the cannuIa from slipping out of the artery during the perfusion. The arteria1 tubing is then attached and pumping is performed by quickIy and gentIy squeezing and releasing the bulb. The vascuIar pressure is determined and sensed by the fingers on the buIb. FIow rates of 50 to 75 ct./minute and Iow pressures are desirabIe. If both sides of the neck are to be perfused simuItaneousIy, it is better to use two separate pumps rather than a single pump connected to the two sets of vesseIs by “Y” tube connections. In using the Iatter method variations in the resistance within the vesseIs wiI1 favor the fIow of perfusate to one side over the other. Determination of the fIow rate may be made during the perfusion as foIIows : The Aow meter and its tubing are fiIIed with heparinized saIine or bIood. One arm of the meter’s tubing is attached to the adapter on the venous Iine. Hemostats are then pIaced on both arms of the tubing. At a given signa the cIamp on the arm attached to the adapter is released and pIaced on the venous catheter between the adapter and the patient. The rate of pumping is maintained as before and continued for five, ten, twenty or thirty seconds at the end of which time the cIamp on the venous tubing is reIeased and pIaced back on the meter’s tubing. The flow rate in miIIiIiters per minute is then read directIy off the caIibrated reservoir. The chemotherapeutic agent is administered through the adapter on the venous Iine which, being proxima1 to the pump, prevents streaming. When HN, is being used the dose is divided into aliquots given at five minute intervaIs. Other agents, being less vesicant and of Ionger duration, may be given as a single injection. Prior to the injection of the chemotherapeutic agent 2 cc. of 5 per cent fluorescein may be injected into the arterial adapter. The room is darkened and the face iIIuminated with an

Simplified TABLE REStJI.TS

OF

HEAD

PERFUSIONS AND

Perfusion TABLE

MALIGNANCIES

ACCORDING

THERAPEUTIC

for RegionaI

II

OF

NECK

Technic

TO

AGENT

OF

THE

111

KESLXTS OF I’EKFL’SIONS OF MALIGNANCIES AXD

CHEMO-

NECK

.ACCORDING

TO

OF THE

HEAD

DIAGNOSIS

USED

Objective TUIIIIX Response

Adjuvant Response

Positive

E. 73.

TSPA...... CB 1348..

AT-DNA......’ Methotrexatc. Total.

/p

/-

-/

Total

Total

Drug

Dihydro HNt..

/-. ~-i-I

Diagnosis

Negative

NO Recurrence

I I 0 0

4

7

:

0

2

2

23

2

”

o

.’

0

o

:

Ir

8

~

Sqllnmous ccl1

14 4 2 I I I

:’

~

Recurrence

ultraviolet lamp. The distribution of the fIuorescence corresponds with the area oerfused. Doing this before injecting the drug permits changes in vessel occIusions necessarv to improv; upon the localization of the therapy. Tracer substances may be used to determine the rate of escape of blood from the isoIated circuit. RISA@ (radioactive iodinated serum albumin) was used in this studv. Control samples ‘are drawn from the arm vein and the pump circuit prior to the injection of 50 PC. of RISA into the arteria1 adapter. Specimens of blood are taken at oeriodic intervals from the arm and the pump and counted. The Ieakage rate, as determined in our experience, suggests that in the average case at the end of fifteen minutes sufficient admixture has occurred to justify the termination of the perfusion. As significant isolation does not persist to the termination of the perfusion we have not drained the blood, nor flushed the circuit with fresh blood, to reduce the amount of agent left in the oatient. At ihe termination of the perfusion the cannulas are removed. The vessels are repaired or ligated. In many instances a polyethylene catheter of small diameter (PE 20) is left in the arterv for subseauent intra-arteria1 infusion of drugs. Technic D&ng Radical Neck Dissection. If the perfusion is to be performed during the course of a neck dissection as an adjuvant to I

I

I

carcinoma Floor of mouth. Tongue. Lip

IL

I

L:1rynx M&~O~l~ Malignant

1

mixed

tumor

of parotid.. Sarcoma of mrninges Total.

0

, I-

0

I1

8

surgery, the neck dissection is carried out in the routine manner. WhiIe the specimen is being swept upward and prior to dividing the externa maxillary vesseIs the external carotid artery and the already divided and usually distended internal jugular vein are cannulated. The perfusion is then performed as described above. At the completion the vessels are repaired or Iigated and the dissection proceeds in a routine manner. The indications for tracheostomy are the same as those for the surgical procedure performed. SeveraI types of chemoDrug Selection. therapeutic agents were studied in this series of cases [6]. (Table II.) Dihydro E-73, an anti. . biotic similar in structure to actidione, was used in fourteen cases [7]. Three alkylating nitrogen mustard, triethylene thioagents, phosphoramide and Leukeran,@ were empIoycd in seven cases. Methotrexate, an antimetaboIite, was used in one case, and altered tumor DNA, obtained from the patient’s tumor, was used in one case. RESULTS

Twenty-three head and neck perfusions were performed on twenty-one patients between January 1959 and January 1961. (Table III.) Nineteen were performed for squamous cell carcinoma, of which twelve were of the Aoor of the mouth, three of the tongue, three of the Iower Iip and one of the larynx. One patient each with melanoma, malignant mixed tumor of the parotid and sarcoma of the meninges

GoIomb,

HaII, Cox, SchetIin,

Gumport

TABLE RESULTS OF PERFUSIONS OF MALIGNANCIES

IV

OF THE HEAD AND NECK

USE

OF

I

Positive

TO USE OR LACK

OF

Adjuvant

Nonoxygenated

II

II

I

I

Negative

ACCORDING

OXYGEN

Oxygenated Diagnosis

and Wright

Total

PClsitive

Negative

Adjuvant

TotaI

_ Squamous ceII carcinoma

......... Adenocarcinoma................. Melanoma ...................... Sarcoma ........................

2

5

8

3

14

0

2

2

0

0

0

0

0

0

I

0

I

0

0

0

0

0

I

4

7

9

3

16

___-

TotaI ........................

1

2

__2

were aIso treated. Of these, four were performed adjuvant to definitive surgica1 procedures. The remaining patients a11 had advanced maIignant disease considered to be incurabIe. Of the technic empIoyed fifteen were uniIateraI, nine bilateral; seven were oxygenated, sixteen nonoxygenated. (Table IV.) The results in those cases in which perfusion was performed for paIIiation are based upon evidence of marked gross tumor necrosis foIIowing perfusion. Moderate objective regressions, cessation of tumor growth or reIief of pain without marked gross necrosis were not considered positive responses. In ten of fifteen cases of squamous ceI1 carcinoma the response was positive as in one case of melanoma. No response was observed in one case in which perfusion was performed twice for a parotid maIignancy and in another for sarcoma of the (TabIe III.) Of the drugs used meninges. dihydro E-73 was effective in seven of ten cases of squamous ceI1 Iesions, HN, was effective in two of three (TabIe v) and Leukeran was effective in the one case in which it was tried. In those cases in which perfusion was performed with an oxygenated circuit, two of six patients responded whereas in those in which the nonoxygenated technic was empIoyed, nine of thirteen patients responded. (TabIe IV.) The visibIe tumor changes became apparent within five days of perfusion and were characterized by a darkening and drying of the surface of the tumor. This was foIIowed by necrosis, liquefaction and slough. Microscopic changes ranging from aIterations in nucleocytopIasmic ratio and increased granularity to

frank necrosis, were observed. The amount of tumor destruction was occasionaIIy extensive, but in a11 cases residua1 tumor has been grossIy visibIe or demonstrabIe by biopsy at the compIetion of the separation of sIough. The duration of antitumor effect in this series averaged thirty-four days. The response to dihydro E-73 averaged twenty-one days whiIe the two patients who responded to HN2 remained free of gross disease for ninety to 120 days before tumor regrowth was noted. The detaiIs of dosage, response and complications are tabuIated in Table v. Four perfusions were performed at the time of extirpative surgery as supplementary or adjunctive therapeutic measures. One patient (Case VIII) with a carcinoma of the floor of the mouth who was perfused with 3.5 mg. dihydro E-73 at the time of a singIe stage hemimandibulectomy and radica1 neck dissection is aIive and free of recurrent carcinoma twenty months later. Another patient (Case XXIII) with a carcinoma of the tongue was perfused with 56 mg. dihydro E-73 during a singIe stage hemigtossectomy, partial mandibulectomy and radical neck dissection. There is no recurrence one month Iater. Another patient (Case XVI) whose Iarge carcinoma of the posterior third of the tongue was incompIeteIy excised during a singIe stage hemigIossectomy and radical neck dissection was perfused at the time with 28 mg. dihydro E-73. A IocaI recurrence appeared seven months later. The fourth patient, an eighty-four year oId man (Case XI) died of a cerebrovascuIar accident, which occurred the day foIIowing a perfusion with 20 mg. HN, combined with a hemigIossectomy, hemimandibuIectomy and radical neck dissection.

-

SimpIified

Technic

-

-

for Regional TABLE

Patient,

Age

(Y*.)

H. K., 74, F, 362 L. L’., 33. F, 363

M. K., 75, M. 3-l II. K., 74, F, 362 A.

L..,

63,

M, 384 J.

C.. M,

F.

60,

386

K.,

72,

F, 365

c.,

F.

46.

Dihydro E--3

10.4 mg. 0.2 “lg./kg.

1Llnlignant mixed t”nlor parotid 1ticl:~nomn of orbit, right

IUnilateral. genated

oxy-

Dihydro E--3

7.84 mg. 0.1 “Ig./ky.

IUnilateral, oxygenated

non-

l-SPA

45 mg. mg./kg.

;quwlous cell carcinoma of floor of mouth IMalignant mixed t”“~or parotid ‘ ;q”~“lous cell carcinoma of floor of mouth ! jquamous cell carcinoma of lXY”X jqoxmous cell carcinoma of flocx of moutl, I

Bilateral, oxygenatcd

“on-

Dihydro E--3

30 mg. 0.6 mg./kg.

Unilateral. genated

oxy

At-DNA

I

IBilateral, genated

oxy

TSI'A

Unilateral, gcnated

oxy

CB

Bilateral, oxygenated

“on.

Unilateral, oxygenated Unilateral,

“0”.

‘

s.,

1:

I

J.

74,

D., 35, F. 475

J.

M., M.

:

451

*Tumor regression.

I ,348

-9 mg./kg. Dihydra

I.5 mg. ,.I mg./kg.

E-73

oxy

Methotrexatc: HNr

~5 w. .9 mg./kg. !O “lg. j.37 mg./kg.

oxy

HNz

LO “lg. 1.3 mg./kp.

Bilateral, genated

HN:!

Lb.4 mg. 3.4 mg./kg.

oxy-

,q1,amous ccl carcinoma 0 floor of moutl

Bilateral, oxygenated

non-

1Dihydro E-73

56 mg. 3.9 mR./kr.

Sq?lnm”“s carcinnma lip

Unilateral. oxygenated

“on-

Dihydro E-73

56 mg. I mg./hg.

Unilateral, oxygenated

“on-

Dihydro E--3

28 “lg. 0.56 mg./kg

Bilateral, oxygenated

“on-

Dihydro

98 mg. 1.7 mg./kg.

Unilateral, oxygenated

no”.

ccl 0

64

Squamous carcinoma lip

responsr:

ccl 0

‘I

E-73

Dihydro

t”mor

regression;

98 mg. 1.6 mg./i.g.

E-73

I

2. objective

A large cervical Iy”lph 1”odc 011 thr perbecame smaller nnd softer; fused side died of disrwc 2)i ,110. I:~tcr Perfusion combined with II single stage hcmim~ndil,ulect~,my x”cl radical neck without diswctio” :Idded Inorhidity: patient is alive and well 1” “~(1. later No obicctive t”“1or respo”w

20 mg.

14 mg. >.*5 mg./kg

496

-

‘j mg. .4 mg /Ig.

Dihydro E-73

,I

N. B.. 61 M. 511

mg.

no”

E

50,

4.6

, 18“w/kg.

Bilateral, oxygenated

74,

S.,

M,

I

jquamous cell carcinoma 01 RwJr of mouth

M. 4b7

I

Marked reduction in tunlor sire lasting 3 mo. and alleviation of pain; palliation wac remarkable in that the mandible. cervical “odes and skin wcrc inrolved with tomor; &cd following inf”sio” therapy 4 mo. later Perfusion combined with a singI? stapr hemimandibulectorny, hemiplosaectomy and radical neck dissrction: hc died rlf :i

‘I


451

E. M.,

F.

of

iquamous cell carcinoma of Ho,~r of moutt

421

M,

(mg./kg.)

“on-

M.,84. M, 422

M.

DOX (Total)

Drug

IUnilateral, oxygenated

C. v., 55. hl, 413

I..

v

ciquamous cell carci”om:l 01 lloor of mouth

D. T.,8, ‘ jarco”m meninges F,408

J.

Type of Perfusion

Diagnosis

Sex and C. K. No.

Perfusion

-

I

1

I

I, intlvmediate

843

response

iadjuvnnt

to drcnitivr

surgery);

o, “O ohirctiv<,

t”m”r

Golomb,

HaII, Cox, SchetIin, TABLE

v

Gumport

and Wright

(Continued)

Patient,

C%SC Age (yr.) Sex and NO. c. R. No.

Type of Perfusion

Diagnosis

Drug

_ XIX

.I. S., 55. M,

xx

XXI

510

w. Y.. 53. M, 525

P. D.. 53, M, 5x2

Type of TLIIWX

DO%

(Total)

Remarks

1Response _-

Necrosis and shrinking of tumor noted; infusion of dihydro E-73 WBS begun on 2nd postoperative day, did well until 8th postoperative day when he suddenIy and unexpectedly died. No cause WBS found at autopsy Response was good, but regrowth began 3 weeks after perfusion

squamous cell carcinoma of tongue

BiIateraI, nonoxygenated

Dihydro

Squamous cell carcinoma of Aoor of mouth squamous cell carcinoma of Aoor of mouth

Unilateral. non. oxygenated

Dihydro

Unilateral. nonoxygenated

HN1

28 mg. 0.5 mg./kg.

Complete necrosis of a11 visible tumor: response was a?sociated with necrosis and edema of face, cheek, neck and pnralysis of seventh nerve; recurrence noted 5 mo.

Di hydra

56 mg. I mg./kg.

Partial necrosis of large tumor regrowth noted 4 weeks later

56 w. 0.7 mg./kg.

Perfused at time of hemiglossectomy, partial msndibulectomy and radica1 neck dissection; no recurrence I mo. later

E-73

E-73

54 mg. I.I mg./kg.

54 w. I mg./kg.

her XXII

xxm

J. V., 46. M. 557. A.

M., 66,

M,

619

ceil Uniiateral, non. oxygenated carcinoma 01 mouth UniIateraI. non. Squamous cell oxygenated carcinoma ol

Squ%mous

E-73

Dihydrc E-73

COMMENTS

The clinica resuIts in this series of cases must be interpreted according to the disease treated and the drugs used. Most of the patients with squamous ceI1 carcinomas were treated with dihydro E-73, a new drug undergoing its initial chnical trial by perfusion [8]. While three of ten patients did not respond to treatment, two of them died in the immediate postoperative period due to excessive doses and the third had responded we11 to a prior perfusion with the same drug. NevertheIess, the duration of effect was shorter with dihydro E-73 than with nitrogen mustard. As its toxic manifestations do not incIude IocaI tissue necrosis or bone marrow depression, it may prove to be a usefu1 drug in combination with one of the aIkyIating agents. The negative responses to the patients treated with altered tumor DNA and methotrexate can be attributed to the choice of agents used. WhiIe it is generaIIy accepted that the benefIcia1 effects of regional perfusion depend to a Iarge extent upon the degree of isoIation achieved, experience with perfusions of the head and neck suggest that compIete or nearly compIete isoIation is not necessary to achieve resuIts. The degree of isoIation ranges from Iess than 25 per cent [4] to as high as 85 per cent [9]. The significance of these observations with radioactive tagged albumin may not be directly applicable to the rate of escape of the

resuIted;

-

chemotherapeutic agents being perfused. For, despite the apparently high rate of leakage intenseIy IocaIized effects are noted within the perfused zone when aIkyIating agents are used. The cIinica1 response in some patients with poorIy isoIated tumors may aIso be due to the high drug concentrations achieved, aIbeit for brief periods. The rate of drug leakage from the isoIated circuit in a variety of area perfusions is ihustrated in Figure 2. These curves are derived from our studies of the escape of RISA from these circuits. It appears that rather high concentrations are attainabIe in the head area, despite a rapid rate of drug Ieakage. This high initia1 drug concentration may contribute to the effects achieved. The use of the nonoxygenated circuit is predicated upon the Iimited importance of an oxygenator in a partiaIIy isoIated area. In regionaI perfusions, such as the head and neck, in which isolation is poor, the increase in oxygen tension with the oxygenators currentIy empIoyed is onIy sIight. Furthermore, the necessity of oxygenating the bIood perfusate is in doubt. Its use is based upon the enhanced effect of ionizing irradiation attained by increasing the tissue oxygen tension and the assumption that a simiIar enhancement may occur with chemotherapeutic drugs. Studies with transpIanted mouse tumors treated with HN2 in the presence of increased tissue oxygen tension tend to support this contention [IO]. But, Ausman and Aust [II] using the

Simplified

j MGM iGM

Technic

\

\

for RegionaI

Perfusion

HEAD

/ 100 cc

TOTAL

DOSE

CONCENTRATION

k-r---TIME

EPUILIBRIUM 1

IO IN MINUTES

AT 15

AFTER

7

20 INJECTION

25 OF

30

DRUG

FIG. 2. These concentration curves are based upon the rate of Ieakage of RISA from the isoIated circuits during perfusion. The curves are averages of our experience in perfusions of the head (unilatera1 and bilatera1 external carotidinternal jugular circuits), brain (unilateral interna carotid-internal juguIar circuits), peIvis, arm and Ieg. The drug concentrations illustrated are for a hypothetic patient with a whoIe blood voIume of 5 L. The concentration at equilibrium is that which would be achieved if the quantity of drug used in the perfusate had been given to the patient intravenously, or if compIete mixing between the isolated and systemic circuIations takes pIace. With the exception of nitrogen mustard, actinomycin D and phenylalanine mustard which are damaging to blood vessels in high concentrations, the doses empIoyed are the maximum which it is estimated the patient could receive systemicaIIy, at a single instance, without producing severe systemic toxicity. The concentration of drug delivered to a tumor-bearing area at the beginning of a perfusion is in the order of ten times for the pelvis, eight times for the leg, ten times for the arm, twentv-five times for the brain and thirtv-three times for the head as could ‘be safely deIivered by any single systemic dose.

fusion eliminates severa of the factors which tend to complicate the procedure principahy: the services of a trained professional and technical team, bIood for priming and the complex oxygenator, heat exchanger and of pump, Iengths of sterile tubing and connectors. This simplified approach maintains the important primary principIe of regional perfusion which is recirculation of a high concentration of drug through the tumor-bearing area. It permits the use of the majority of effective chemotherapeutic agents which, having no antidotes, wouId not be suitable for infusion therapy. As the dissection for this perfusion is essentially the same as for the installation of indwelling arterial catheters, there is very little extra surgery added. While we have not made it our usua1 practice to follow perfusion with infusion therapy, in order to fuIIy evaluate the effects of the former method, it seems logical that to achieve the best resuIts from chemotherapy a combination of modaIities might be

perfused hind limb of a dog noted greater uptake of nitrogen mustard by the experimental leg maintained at 70 to 85 per cent oxygen saturation than the contro1 leg at IOO per cent oxygen saturation. The enhancement of antitumor effect by an oxygenated circuit for perfusions of the head has not been supported by the resuIts of this study. In 33 per cent of the oxygenated cases a positive antitumor response was achieved while in the nonoxygenated series 70 per cent responded. The brief duration of perfusions of the head and neck and the high degree of colIatera1 bIood supply make the use of an oxygenated circuit unnecessary to maintain tissue viability. There is a tendency among some surgical chemotherapists to abandon perfusion in favor of infusion therapy for tumors of the head and neck [4,12]. One of the motivating factors is the duration and complexity of standard perfusion procedures. The nonoxygenated per845

GoIomb,

HaII, Cox, SchetIin,

used. FolIowing the perfusion with an aIkyIating agent or antibiotic, fine catheters can be Ieft in the appropriate vesseIs for the subsequent infusion treatment with antimetaboIitemetaboIite therapy described by SuIIivan [13]. The concomitant use of genera1 hypothermia and hyperthermia in the isoIated circuit as advocated by Singleton [14] in perfusion of the abdomina1 organs might Iend itseIf to perfusions of the head and neck area. With the use of rapid acting aIkyIating agents systemic toxicity can be reduced; however, the limiting factor of dosage in this area is usually IocaI tissue toxicity. CONCLUSIONS

A simpIified, nonoxygenated method for performing regional perfusions and its applicabiIity in paIIiative and adjuvant perfusions of the head and neck area is presented. The results of a cIinica1 experience with twenty-three cases of perfusion therapy for maIignancies of the head and neck is presented. SeveraI chemotherapeutic agents are effective in obtaining seIective tumor necrosis in this area. The drugs found to be of vaIue in this series are: dihydro E-73, nitrogen mustard, triethyIene thiophosphoramide and Leukeran. REFERENCES

I. RYAN, R. F., KREMENTZ, E. T., CREECH, O., JR., WINBLAD, J. N., CHAMBLEE, W. and CHEEK, H. SeIected perfusion of isoIated viscera with chemotherapeutic agents using an extracorporea1 circuit. S. Forum, 8: 158, 1957. 2. WOODHALL, B., PICKRELL, K. L., GEORGIADE, N. G., MAHALEY, M. S., JR. and DUKES, H. T. Effect of hyperthermia upon cancer chemotherapy: appIication to external cancers of head and face structures. Ann. Surg., 151: 750, 1960.

846

Gumport

and Wright

3. AUST, J. B., AUSMAN, R., GRIFFEN, W. and ZIMMERMAN, B. The use of isolated perfusion chemotherapy techniques in head and neck cancer. Presented at Fifth AnnuaI Meeting of the Society of Head and Neck Surgeons, Washington, D. C., March 31, 1959. To be published. 4. JESSE, R. H. Technique of perfusion and infusion of the head and neck. Cancer Cbemotberap. Rep., IO: 21, 1960. 5. CREECH, O., JR., KREMENTZ, E. T., RYAN, R. F. and WINBLAD, J. N. Chemotherapy of cancer: regional perfusion utilizing extracorporea1 circuit. Ann. Surg., 148: 616, 1958. 6. GOLOMB, F. M. Selection of drugs for perfusion. Cancer Cbemotberap. Rep., IO: 53, rg6o. 7. WRIGHT, J. C., GUMPORT, S. L. and GOLOMB, F. M. A drug with antitumor effects in man. Preliminary cIinicaI report. Cancer Cbemotberap. Rep., 8: 7, 1960. 8. GOLOMB, F. M., WRIGHT, J. C., COBB, J. P., GZJMPORT, S. L., POSTEL, A. and SAFADI, D. The chemotherapy of human soIid tumors by perfusion techniques. Proc. Am. A. Cancer Res., 3: 92, 1960. g. AUST, J. B. Discussion. Cancer Cbemotberap. Rep., IO: 22, rg6o. IO. KREMENTZ, E. T., HARLIN, R. and KNUDSON, L. The enhancement of chemotherapy by increased tissue oxygen tension. Cancer Cbemotberap. Rep., IO: 125, 1960. II. AUSMAN, R. K., GEMMILL, S. J., JERNBERG, E. J. and AUST, J. B. Effect of oxygen on the uptake of nitrogen mustard in isoIated perfusion. Proc. Am. A. Cancer Res., 3: 92, 1960. 12. HERTER, F. P., MARKOWITZ, A., PRICE, J. B., WEISSMAN, S. G. and DEMETZ, A. Cancer chemotherapy using antimetaboIite-metaboIite combinations. S. Forum, I I : 52, 1960. 13. SULLIVAN, R. D., MILLER, E. and SIKES, M. P. AntimetaboIite-metaboIite combination cancer chemotherapy. Effects of intra-arteria1 Methotrexate-intramuscular citrovorum factor therapy in human cancer. Cancer, 12: 1248, 1959. 14. SHINGLETON, W. and PARKER, R. T. AbdominaI perfusion with hypothermia and hyperthermia. Cancer Cbemotberapy. Rep., IO: 35, 1960.