haematopoietic growthfactors andlung - thoraxshort bolus intravenous injections, subcutan-_ eous...

Thorax 1992;47:1 19-126

New perspectives in lung cancer 4 Series editor: S G Spiro

Haematopoietic growth factors and lung cancer

treatment

N Thatcher

Normal haematopoiesis calls for a complicatedbut integrated process of proliferation anddifferentiation, with about 7 x 109 granu-locytes and 1 x 1010 erythrocytes replacedhourly. The haematopoietic colony stimulat-ing factors (CSFs) are glycoproteins that werefound to stimulate and promote the prolifera-tion of granulocyte-monocyte progenitor cellson semi-solid media in clonogenic assays(table 1, fig 1). The diverse but orderedinteractions of colony stimulating factors, tar-get cells, and stroma are responsible for thecomplex process of haematopoiesis. Thecolony stimulating factors have effects otherthan proliferation and differentiation, includ-ing maintenance of cell viability, membraneintegrity, and functional stimulation ofmature cells-for example, granulocytephagocytosis and superoxide production(table 2).'-'The growth factors have very different

structures and specific receptors exist on targetcells, though a target cell may possess manymore than one type of colony stimulatingfactor receptor. Growth factors are alsounusual in having very high biological activityand in being able to increase rapidly givenappropriate stimuli. The production ofgrowth factors from endothelial and stromalcells, fibroblasts, lymphocytes, and macro-phages is controlled by a network of inter-actions between the various cell types andexternal stimuli, such as foreign antigen andendotoxin. In addition, we are becomingincreasingly aware of important synergisticintereactions between growth factors, whichagain emphasise the complexity of the systemand the potential difficulties of full clinicalexploitation. 1-3

Various factors other than the colonystimulating factors, such as erythropoietin andthe interleukins, cause proliferation and dif-ferentiation of cells, the latter on B and Tlymphocytes. Some, such as interleukin-3(IL-3) and IL-6, also have an effect onmyeloid cells (table 1, fig 1). Growth factorssuch as IL-3 that act on multipotent progeni-tor cells of early lineage give rise to a range ofmature cell types, including erythrocytesplatelets, monocytes, and the various granu-locytes. Granulocyte-macrophage colonystimulating factor (GM-CSF) stimulatesproduction of granulocytes and monocytesand increases the levels of eosinophils andlymphocytes, and, in some cases, platelets anderythrocytes. G-CSF is much more lineagerestricted and acts specifically on neutrophilgranulocytes (table 1, fig 1).The genes coding for IL-3, GM-CSF, and

monocyte (M) CSF are localised to a smallarea of chromosome 5, whereas the gene forG-CSF resides on chromosome 17 (table 1).Recombinant DNA technology has enabledthese factors to be further examined both invitro and in early clinical studies. Data arenow available on erythropoietin, G-CSF, andGM-CSF and preliminary results will soon beavailable for IL-3, IL-4, and IL-6.

Clinical applications of growth factors inlung cancerHaematopoietic growth factors are now thecentre of feverish clinical activity in a widerange of malignant and non-malignant dis-eases, including solid tumours, haematopoieticmalignancies, myelodysplastic syndromes,aplastic anaemia, AIDs, and the idiopathicneutropenias (table 3).'-'

Table I Human haematopoietic growth factors of current clinical interest*

Molecular Location ofName Abbreviation weight (daltons) gene

Erythropoietin Epo 39 000 7q 11-22Granulocyte colony stimulating G-CSF 20 000 17q 11-2-21

factorGranulocyte-macrophage colony GM-CSF 18-30 000 5q 23-31

stimulating factorMacrophage colony M-CSF (CSF-1) 70-90 000,

stimulating factor 45-50 000 5q 33-1Multipotential colony Multi-CSF (IL-3) 15-30 000 5q 23-31

stimulating factorInterleukin-4 IL-4 16-20 000 5q 31Interleukin-6 IL-6 19-21 000 7p 15

*Modifed from Metcalf2.

ERYTHROPOIETIN

Erythropoietin has produced dramaticimprovement in the anaemia and quality of lifeof patients with end stage renal failure,provided that sufficient iron stores are present.Patients with malignant tumours, includinglung cancer, often have normochromic-nor-mocytic anaemia with a block in iron transferfrom stores to the erythroid precursor cell.They may also have anaemia due to blood lossor marrow infiltration and, interestingly, theerythropoietin level tends to be lower inpatients with malignant disease (including lung

Cancer ResearchCampaignDepartment ofMedical Oncology,Christie Hospital andHolt Radium Institute,Manchester M20 9BXN ThatcherRequests for reprints to:Dr Thatcher

119

on August 21, 2021 by guest. P

rotected by copyright.http://thorax.bm

j.com/

Thorax: first published as 10.1136/thx.47.2.119 on 1 F

ebruary 1992. Dow

nloaded from

http://thorax.bmj.com/

Thatcher

Figure I Targets ofhaematopoietic growthfactors. IL-3-interleukin-3;CSF-colony stimulatingfactor; M-macrophage;G-granulocyte.

Multipotentstem cells

Lineage restrictedprogenitor cells

More matureprecursor cells

IL-3(Multi-CSF)

Mature cellsproduced

ErythrocytesNeutrophilsBasophilsMonocytesMegakaryocytesEosinophils? B cells

EosinophilsMonocytes/macrophagesNeutrophils

Monocytes/macrophages

G-CSF - Neutrophils

Erythropoietin . Erythrocytes(platelets?)

cancer) than in patients with anaemia due toother causes but of similar severity.45 Theerythropoietin response was also found to bedecreased in patients receiving chemotherapy,though this was not due particularly to the useof nephrotoxic drugs such as cisplatin.5A randomised study, reported so far only in

preliminary form, examined the effect of ery-thropoietin at two doses and compared thiswith outcome in a control group.6 All patientshad small cell lung cancer and were treated withan intensive carboplatin-ifosamide-etoposideregimen. Suppression of bone marrow madefrequent blood and platelet transfusions man-datory. Erythropoietin (150 or 300 IU/kg threetimes a week subcutaneously) resulted in trans-fusion of significantly less blood in both groupsreceiving erythropoietin than in the controlgroup and in a trend towards fewer platelettransfusions. The latter trend strengthens the

Table 2 In vitro actions of myeloid colony stimulating factors*

1

23

4

Maintain survival at all stages of development of granulocytes and monocytesPresence required to induce cell division; concentration determines length of cell cycleCommit bipotential granulocyte-macrophage precursor cells to enter granulocytic ormonocytic lineageStimulate functional activity ofmature polymorphs and monocyte-macrophages;effects on chemotaxis, expression ofmembrane antigens, phagocytosis, superoxideproduction, killing ofmicroorganisms and tumour cells, and production ofbiologically active agents (for example, interferons, tumour necrosis factor,prostaglandins)

*Modified from Steward et al.'

Table 3 Potential clinical applications of myeloid colony stimulatingfactors*

1 Treatment of bone marrow failure:(i) idiopathic

(ii) neoplastic(iii) iatrogenic

2 Augment rate of recovery after bone marrow transplantation3 Reduce duration or degree (or both) of leucopenia following chemotherapy4 Increase granulocyte number and function (for example, in patients with AIDS)5 Treatment of leukaemia-altering the rates of cell reproducton and differentiation6 Treatment of myelodysplasia-increase normal differentiation and reduce blast

population7 Treatment of established bacterial and fungal infections8 Improve host defence against potential infection after major trauma-for example,

burns

*Modified from Steward et al.'

controversial experimental observation ofmegakaryocyte precursor stimulation by ery-

thropoietin. There may therefore be a case forthe use of erythropoietin to reduce bloodtransfusions in selected patients with cancer.

GRANULOCYTE COLONY STIMULATING FACTORAND CONVENTIONAL CHEMOTHERAPYThe first clinical study of a colony stimulatingfactor, G-GSF, was conducted in Manchesterin patients with small cell lung cancer receivingchemotherapy.7 This and subsequent studiesexamined the effect of G-CSF after conven-tional dose chemotherapy in preventing druginduced neutropenia. In the Manchester studypatients received up to six cycles of treatmentwith doxorubicin, ifosfamide, and etoposideand were randomised to receive G-CSF in odd(Nos 1, 3, 5) or even (2, 4, 6) cycles. The G-CSF was given by continuous 14 day intraven-ous infusion through an ambulatory pump,starting the day after chemotherapy. In addi-tion, the dose-response relationship (from 1 to40 ug/kg/day) was examined over five daysbefore the first course of chemotherapy. In thisstudy the maximum response to G-CSFoccurred with the 10 gg/kg/day dose and G-CSF was extremely effective in reducing severeneutropenia (defined as less than 1 x 109/lneutrophils-a level considered to be critical asbelow this patients' vulnerability to lifethreatening infections and death from sep-ticaemia is greatly increased). The duration ofneutropenia was substantially reduced (median80%) with G-CSF (fig 2) and normal or abovenormal neutrophil counts were obtained withintwo weeks of chemotherapy. Of particularimportance was the observation that all six lifethreatening infections occurred after cycles ofchemotherapy without G-CSF and no severeinfection occurred after cycles in whichpatients were protected with G-CSF. Thesevere infections resulted in 30 extra days inhospital for intravenous antibiotic treatmentand other supportive measures.7Other studies have confirmed these observa-

tions, particularly the specific effect on neutro-phils and the fact that no toxic effects, or only

GM-CSF

M-CSF

.

b

120



j.com/


ebruary 1992. Dow

nloaded from


Haematopoietic growth factors and lung cancer treatment

IVG-CSF antibiotics

CTCT

G-CSF

CT CT

0 20 40

a

BT CT G-CSF

60 80Days

Figure 2 Haematological response to granulocyte colony stimulatingfacshowing changes before chemotherapy (CT) and afterfour cycles. IV-inBT-blood transfusion. The shaded area indicates the total area of absoluneutropenia.

very minor side effects, occurredGabrilove investigated patients Mtransitional cell bladder cancdoxorubicin, cisplatin, vinblmethotrexate chemotherapy whcintravenous infusion of G-CSF 4utes.8 They found a reduction in tdays of neutropenia (less than 1 >

phils/1), and all patients were ableplanned chemotherapy on day I

with only 29% in courses with Ifrom G-CSF. Interestingly, the]reduction in mucositis. These isuggested the possibility ofchemotherapy, with courses giverrather than the conventional threeintervals.

In a larger study in 126 patiencell lung cancer reported by (abstract form) patients were gitional doses of cyclophosphamid4and etoposide and randomised tCSF or no G-CSF on days 4-17otreatment cycle.9 G-CSF was giicutaneous bolus of230 jg/m2/daysignificant reduction in the incideition of neutropenia and in thesevere infection as manifestneutropenia. Benefit occurred bocycle of chemotherapy, wheninfection in patients with small ce

receiving chemotherapy apppredominant and in subsequentnumber of days spent in hospital;intensive antibiotic treatment wasin patients treated with G-CS]study is now being perform(European centres with the same cc

drugs. Preliminary results indicatin febrile neutropenia with a proextending through all six cycletherapy, and a substantial reducti(spent in hospital with infections.about 90% ofpatients having G-Cto receive full dose chemother;compared with only 65% of patieplacebo.'°Another group of investigators,

compared the response to G-C

short bolus intravenous injections, subcutan-_ eous bolus, and subcutaneous infusion atCT various doses." 12 G-CSF again reduced the

duration of neutropenia after melphalanHaemoglobin chemotherapy, though patients previously

given chemotherapy or radiotherapy did notPlatelets appear to respond as well to G-CSF as diduntreated patients. When G-CSF was givenWhite cells subcutaneously to 31 patients, including nine'Neutrophils with lung cancer, it was again very well

tolerated and substantially reduced theneutropenia, even when G-CSF was startedseveral days after the melphalan. As in otherstudies using G-CSF, there were no changes in

_T_r_r__,the counts of cells other than neutrophils. Adose of 3 ug/kg/day produced similar increasesin neutrophil counts when given by bolus and

ror (G-CSF) by continuous subcutaneous infusion, theitravenous; neutrophils increasing within 24 hours withtte both routes." In a Japanese phase I/II study of

33 patients with primary lung cancer of allwith G-CSF. histological types various chemotherapyrith advanced regimens were used,'3 The optimal dose of G--er receiving CSF in these patients was 100-200 pg/m2 givenlastine, and as a 30 minute intravenous infusion over 14had a short days; 400 pg/m2 was recommended for patients

over 30 min- who had had previous chemotherapy, becausehe number of the bone marrow response would bex IO' neutro- impaired.'3to receive the14, comparedno protection GRANULOCYTE-MACROPHAGE COLONYre was also a STIMULATING FACTOR AND CONVENTIONALnvestigations CHEMOTHERAPY

accelerated In the first published clinical trial withn at two week intravenous GM-CSF patients with AIDS andor four week bone marrow failure showed a dose dependent

increase in circulating leucocytes, most beingits with small mature granulocytes. The subject has beenCrawford (in reviewed recently.'5 In this and other studiesiven conven- peripheral eosionophlia was a feature. Thee, doxorubin, stimulation of all types of granulocytes is ato receive G- feature of GM-CSF, whereas G-CSFfa three week specifically stimulates neutrophil granulocytes.ven as a sub- Combinations of GM-CSF with zidovudiner. There was a are now under investigation in patients withnce and dura- AIDS to improve tolerance to zidovudine andincidence of reduce requirements for antibiotics. In someby febrile patients with myelodysplastic syndromes an

th in the first increase in platelet and reticulocyte counts anddeaths from a reduction in tranfusion requirements have11 lung cancer been observed.'5ear to be In 1989 some phase I studies investigatedcycles. The routes of administration and the GM-CSF

and receiving dose-response relationship in patients with; 40-50% less refractory, advanced solid tumours, many ofF. A similar whom had been treated previously. In anothered in some study from Manchester a dose-response)mbination of relationship was observed with GM-CSF, withte a reduction significant increases in total leucocyte, neutro-tective effect phil, and eosinophil counts.'6 Of particulars of chemo- interest was one patient who had received aon in the time large amount of treatment for liposarcoma, inIn addition, whom a 50% reduction of the tumour occurredSF were able after GM-CSF; the response lasted for sixapy on time months. Seven other patients, with tumoursnts receiving that were progressing, were stabilised.'6 Fur-

ther investigation of the potential anticancer, in Australia, activity of GM-CSF (possibly mediatedSF given by through the macrophage system) are warran-

121



j.com/


ebruary 1992. Dow

nloaded from


Thatcher

ted. Leucocytosis, largely due to an increase ingranulocytes was also reported with subcutan-eous GM-CSF given once daily (3-15 pg/kg/day).'7 ThusGM-CSF is an effective stimulatorof haemotopoiesis with a role in reducingcytopenia after chemotherapy and bonemarrow transplantation.' 1518

GM-CSF was examined by Antman et al in16 adults with sarcoma treated with doxo-rubicin, ifosfamide, and dacarbazine.1' GM-CSF was given in the first cycle and theoutcome compared with that of chemotherapyalone in the second cycle. Chemotherapyinduced neutropenia was reduced with GM-CSF, which was well tolerated in doses up to32 pg/kg a day given by continuous intraven-ous infusion. The number of febrile episodes,however, was similar for the two courses ofchemotherapy, and it is not clear whether thesewere due to infection or to pyrexia arising fromthe GM-CSF. A study of GM-CSF in smallcell lung cancer from Manchester had the samedesign as that of Bronchud et al-that is,randomisation betwen odd and even cycles ofchemotherapy.7 With subcutaneous GM-CSF intravenous antibiotics were required inseven of26 cycles, a proportion similar to the 10out of 30 cycles without GM-CSF. Despite thereduction in neutropenia after chemotherapythe incidence of infection and requirement forintravenous antibiQtics were very similar,indicating the need to examine all these clinicalmeasures in studies of the efficacy ofhaemopoietic growth factors.20 In a Germanrandomised investigation in patients with smallcell lung cancer subcutaneous GM-CSF shor-tened the duration of neutropenia induced bychemotherapy that include doxorubicin andifosfamide, and it reduced the incidence ofinfections requiring antibiotics.2"

PHARMACOLOGY AND TOXICITYThe qualitative effects ofG-CSF andGM-CSFon peripheral blood neutrophils are similar.Within 30 minutes of administration of thecolony stimulating factor there is a transient fallin peripheral blood neutrophils followed by arapid and substantial increase to above baselinelevels about five or six hours after administra-tion. The transient depression in the count maybe caused by margination to endothelial cells.The increase in neutrophil count with G-CSFand the leucocytosis with GM-CSF reflectdemargination, accelerated release of cells fromthe bone marrow, and an increase in cellproliferation. The dose dependent increases incell count may be many times greater than thebaseline count, tending to plateau from twodays onwards.' 11121416-18 For GM-CSF theincrease in white cell count may be phasic,affecting both neutrophils and (to a lesserextent) eosinophils.'6 Retreatment with GM-CSF produced similar changes in the leucocytecounts but with a more rapid increase andsomewhat higher peak counts.'617 After G-CSF or GM-CSF is withdrawn the cell countsdecline rapidly and by 24-48 hours they areback to baseline levels. When intravenous andsubcutaneous administration of G-CSF andGM-CSF were examined the increase in peri-

pheral blood counts appeared somewhatgreater with the subcutaneous route, an impor-tant practical observation.'""1618 The neutro-phils produced in response to G-CSF have anin vitro functional capacity similar to that ofbaseline neutrophils." Phagocytic function asreflected by chemoluminescence is usuallyincreased by G-CSF and less consistently withGM-CSF.14 20

G-CSF has been extremely well tolerated,with only occasional bone pain and some mus-culoskeletal discomfort occurring with higherdoses. The discomfort is mainly in themedullary areas (stemum, jaw, pelvis, back,and limbs), usually lasts for only a few hours,and does not necessarily recur with subsequentdoses. G-CSF has been associated with revers-ible increases in lactate dehydrogenase, alkalinephosphatase, liver transaminases, and uric acidand in some patients a reduction of serumcholesterol.3 1518 More severe side effects havebeen reported with GM-CSF but theseoccurred with the higher doses in phase Istudies. Side effects of GM-CSF also includebone pain and fever in some patients. Acapillary leak syndrome has resulted in fluidretention with pericardial and pleural effusions,fever, arthralgia, hypotension, and renal dys-function, but only with higher doses. Activa-tion of inflammatory cells with overexpressionof adhesion molecules resulting in aggregationof these cells, particularly in the microvascul-ture, may explain in part the capillary leak.

Nevertheless, the increase in leucocytesoccur with doses ofGM-CSF that are tolerableplus an improvement in platelet andreticulocyte counts in some studies.`3 18 In anoccasional patient considerable leucocytosis(>30 x 109/1) has occurred, though this isvery uncommon with the dosages now recom-mended-for example, 5 pg/kg/day G-CSFsubcutaneously for 14 days or until the countshave recovered. Such severe leucocytosisresolves rapidly once the colony stimulatingfactor has been discontinued.There are theoretical drawbacks to the use of

growth factors, including diversion ofhaematopoiesis to a specific cell type withreductions in other cells lines, toxin produc-tion, and marrow hypoplasia after "badlytimed" chemotherapy; in addition, colonystimulating factors are not recommended in the24 hours before or after conventional chemo-therapy. There is also the potential for trans-formation with increased growth rate ofmalignant cells. This is potentially importantas receptors to colony stimulating factors arecapable, in some cases, of modulating cellproliferation in lung cancer cell lines.2224Similar receptors in fresh biopsy material fromsmall cell lung cancers (J Hampson, unpub-lished data) have not, however, been identifiedby our group.With the growth factors currently examined

there has been no evidence of late marrowfailure due to "marrow exhaustion" and noevidence of the type of wasting illness seen inrodents undergoing long term administrationof GM-CSF, thought to be due to excessivemacrophage activation. Neutralising anti-

122



j.com/


ebruary 1992. Dow

nloaded from


Haematopoietic growthfactors and lung cancer treatment

bodies to the growth factors have not beendetected and there have been no reports ofunexpected exacerbation of malignant dis-ease.-3 9 15 18

G-CSF AND GM-CSF FOR INTENSIFICATION OF

CHEMOTHERAPYThe initial studies of G-CSF and GM-CSFreporting a reduction in neutropenic infectionsinduced by chemotherapy and in antibioticrequirements and a low incidence of side effectsled to the examination of growth factors in thecontext ofdose intensification in chemotherapy.The issue is important as there is increasingevidence that dose intensification may beassociated with improved response andsurvival in patients with various solid tumours,including lung cancer.25

Accelerated chemotherapyG-CSF has been used in patients with ad-vanced breast and ovarian cancer to facilitateescalation ofthe dose and rate ofadministrationof doxorubicin. Doxorubicin could be givensafely only at doses of 75 mg/m2 every threeweeks without G-CSF but 125 and 150 mg/m2could be administered every two weeks with G-CSF, an increase in dose intensity of up tosixfold.26 The higher dose rates were associatedwith a much improved rate of complete andpartial response. All patients treated with thetwo higher doses of 125 and 150 mg/M2 respon-ded, though non-myelosuppressive toxicity,particularly epithelial damage, then becamedose limiting.26 The study emphasised that thechoice of the cytotoxic drugs is critical andmust take into account the potential for severenon-myelosuppressive toxicity at doses higherthan the conventional ones.

In a small study by Ardizzoni, in whichaccelerated chemotherapy with cyclophos-phamide, doxorubicin, and vincristine alter-nating with cisplatin and etoposide was givenwithout dose modification, a twofold increase indose intensity was achieved in patients withsmall cell lung cancer by giving GM-CSF.29 Inthe German study in which GM-CSF is beingused preliminary results indicate that this mayallow chemotherapy to be given at two weeklyintervals with a schedule of doxorubicin, ifos-famide, and vincristine alternating with cis-platin and etoposide.2' A further investigationis now being conducted in Manchester withintensive chemotherapy-with carboplatin,ifosfamide, and etoposide-in which dosagesare not modified because of previous toxicity.This regimen or a variation (alternating carbo-platin with cisplatin) has resulted in two yearsurvival rates of 30% or more in patients withlimited disease who had not had computedtomography and in patients with extensivedisease with no other adverse factors.27 2 In thenew investigation patients are randomised toreceive or not receive G-CSF and wheneverpossible the chemotherapy is given at muchshorter intervals than the 4-6 weeks usuallyrequired for this intensive regimen. The studyshould identify any survival benefits that arisefrom dose intensification withG-CSF and showwhether reduction in neutropenia is accompan-ied by fewer infections and sequelae.

Dose intensive chemotherapy with and withoutbone marrow transplantationRepeated courses of cisplatin, etoposide, andcyclophosphamide have been given in doses forwhich transplantation is usually necessary to 24patients with refractory malignancies (includ-ing some with lung cancer). G-CSF againshortened the duration of severe neutropenia ina dose related fashion and enabled thesepatients to have fewer days of antibiotic treat-ment than a control group, but their stay inhospital was not reduced. The protective effectof G-CSF (given as a 30 minute intravenousinfusion) was sustained with repeated cycles ofchemotherapy.'0 In another study, from Aus-tralia, subcutaneous G-CSF was given topatients receiving high dose busulphan andcyclophosphamide with autologous bonemarrow transplantation. Fifteen patients withnon-myeloid but chemosensitive malignancies,all of whom had been previously treated withchemotherapy and some with radiotherapy,were compared with a historical control groupof 18 patients who had received the same highdose treatment alone. The 15 patients had afaster neutrophil recovery (over 0 5 x 109/1)than the control group (mean day 11 v day 20).When compared with the controls there was afewer days of parenteral antibiotic treatment(11 v 18 days), and a fewer days of parenteralnutrition for severe oral mucositis (10 v 16days). The number of days spent in hospital(23 v 30) didnotdiffer significantly. Doses ofG-CSF were decreased stepwise once the neutro-phil count exceeded 1 x 109/l for three con-secutive days and needed to be reintroduced tomaintain the count in only three of the original15 patients.3'The use of GM-CSF after autologous bone

marrow transplantation in patients with lym-phoid malignancies has been reviewed byAppelbaum."8 Again there was acceleratedrecovery in 15 patients given treatment, notonly of granulocytes but also of platelets by oneto two weeks, with fewer infections and earlierdischarge from hospital, whereas less than 5%of 100 patients recovered by this time in thecontrol group without GM-CSF. GM-CSFhas been used to treat graft failure in patientswho after a period of successful engraftmentlose their grafts. Sustained increases in gran-ulocyte counts occurred within two weeks ofthe start of GM-CSF treatment and this wasassociated with improved survival.'8 Inpatients with refractory breast cancer andmetastatic melanoma given high dose treat-ment with cyclophosphamide, carmustine, andcisplatin followed by autologous bone marrowtransplantation GM-CSF doses unassociatedwith substantial toxicity resulted in acceleratedrecovery of the total white cell count. Thepatients receiving GM-CSF tolerated chemo-therapy with considerably less subjective andobjective evidence of toxicity than the controls,with fewer episodes of septicaemia and lesshepatotoxicity and nephrotoxicity, possibly asa result of fewer infections.32 These and otherstudies suggest that accelerated and high dosechemotherapy, with repeated courses, are pos-sible, and this will reopen the issue of repeatedhigh dose treatment for chemosensitivetumours such as small cell lung cancer.

123



j.com/


ebruary 1992. Dow

nloaded from


124

Peripheral blood stem cell harvesting andintensive chemotherapyHaematopoietic colony stimulating factors alsohave the ability to increase levels of progenitorcells in peripheral blood. The progenitors canbe harvested from the peripheral blood byleucapheresis and used as an adjunct to marrowcells, or possibly in place of them, for recon-stitution after myeloablative treatment.'33-36As an example, after high dose cyclophos-phamide chemotherapy there is a 30 foldincrease in peripheral blood stem cells duringthe rapid recovery phase of the drug inducedpancytopenia. The increase can be augmentedby GM-CSF up to a 1000 fold in somecircumstances.3435 Collection of peripheralblood stem cells by leucapheresis by meansof a cell separator, cryopreservation, and laterreinfusion has substantially acceleratedhaematopoietic reconstitution after high dosechemotherapy and radiotherapy in patientswith advanced malignancy, including small celllung cancer.35 6 Even in patients who have notreceived high dose chemotherapy G-CSF andGM-CSF are capable of increasing progenitorcells of both the myeloid and the erythroidseries.33 The most dramatic result is enhancedgranulocyte recovery but the number ofplatelet transfusions may also be reduced.3537The number ofperipheral blood stem cells maybe so much augmented that the cells can becollected with very few leucaphereses-indeed,the number of cells thought to be needed forreconstitution might be obtained from only500 ml of whole blood after chemotherapy andtreatment with colony stimulating factor.Dose intensification is possible by shortening

the intervals between courses. The results of apreliminary study with peripheral blood stemcells obtained by sequential harvesting andreinfusion after repeated courses of high dosecarboplatin and GM-CSF in ovarian cancerhave been described. Myelosuppression andthrombocytopenia were much less severe, withless need for transfusions, intravenous anti-biotics, and stay in hospital than in patientsreceiving GM-CSF only. The use of the peri-pheral blood stem cells also allowed a substan-tial increase (38%) in dose intensity.38 Thesetechniques may well be applied in pilot studiesof selected patients with small cell lung cancer.

The futureThe clinical use of recombinant humanhaematopoietic growth factors has already ledto important advances in the management ofpatients with cancer. Bone marrow impairmentis the dose limiting complication for manychemotherapeutic agents and amelioration ofthe profound neutropenia and associated lifethreatening infections is now possible. Theoptimal schedules for colony stimulating fac-tors are still unknown and not all studies withG-CSF and GM-CSF have shown a reductionin infection. Our current administration ofthese growth factors is unphysiological and aperiodic pulse delivery of colony stimulatingfactors, given the physiological example of

hormone release, would be perhaps moreappropriate. Nevertheless, colony stimulatingfactors have produced clinical effects duringprofound bone marrow suppression whenrelease of growth factors is likely to be at aphysiological maximum. The possibility ofgiving a fixed dose (the contents ofa single phialof G-CSF) is an attractive proposition andshould be considered, given the data from theclinical dose ranging experiments. This wouldallow outpatient treatment with administrationby the patient or a family member much morefeasible. The recovery of haematopoetic stemcells from peripheral blood is likely to be ofgreat importance in enhancing haematologicalrecovery after myelosuppressive treatment andmay even replace marrow transplantation andits associated difficulties. Much work still needsto be performed to determine the correcttiming for the harvest and to find out how toobtain the optimal number of peripheral bloodstem cells with colony stimulating factors usedalone or in conjunction with chemotherapy.

Combinations of colony stimulating factors,including G-CSF and GM-CSF, with othergrowth factors, such as IL-3, cause synergisticeffects in animal models. The combinations arepotentially capable of restoring all aspects ofhaematopoiesis after marrow suppression morerapidly than a single colony stimulating factor.IL-6 is of particular interest as it has a con-siderable effect on megakaryocytes and maywell diminish thrombocytopenia. Theintroduction of megakaryocytic CSF andthrombopoietin also is eagerly awaited forovercoming thrombocytopenia. Treatmentwith IL-3 followed by GM-CSF produces amanyfold increase in the number of progen-itors within a few days of administration inprimates. Such combinations should improvethe yield of peripheral blood stem cells formarrow rescue after repeated courses of inten-sive treatment in the future. There is also thepossibility of obtaining sufficient numbers ofperipheral blood stem cells by venesection, andthus avoiding the logistic difficulties of leuca-pheresis.

Further clinical exploitation ofcombinationsof growth factors is likely to follow in the verynear future. It would seem reasonable, forexample, to use IL-6 to stimulate primitivestem cells followed by IL-3 to augment multi-potent progenitors and GM-CSF to improvemyelopoiesis. These would be followed byother factors, such as G-CSF or erythropoietinto act on more mature precursor cells. Thistype of schedule might lead to maximumstimulation and more rapid recovery of all celltypes after marrow suppression.

Synthetic materials also help to protectagainst myelosuppression and other forms oftoxicity induced by chemotherapy. EthyolWR2721 was developed as a radioprotectivedrug and is taken up differentially by normalrather than malignant cells. Some clinicalstudies indicate that ethyol can protect patientsagainst myelosuppression, neurotoxicity, andnephrotoxicity associated with cyclophos-phamide and cisplatin, an approach alsoworthy of further investigation." 4

Thatcher



j.com/


ebruary 1992. Dow

nloaded from


Haematopoietic growth factors and lung cancer treatment

Some growth factors, such as GM-CSF, IL-4 and IL-6, modify the cell cycle. GM-CSFshortens the total cycle and the duration of theS phase of bone marrow cells; once it is stoppedthe rate of marrow proliferation fallsdramatically to below the pretreatment rate.Scheduling cytotoxic drugs, with GM-CSF toreduce myelotoxicity is another interestingprospect particularly for drugs that are cellcycle specific. If haematopoietic growth factorsand interleukins are required for growth ofmalignant cells (for example, IL-6 for themyeloma cell) via autocrine or paracrine loops,inhibitors of these factors may be of value.Antibodies to IL-6 may help patients withmyeloma, for example. GM-CSF and otherfactors may have antitumour effects in their ownright by enhancing the activity of myeloid cellsand inducing release of cytokines (such asinterferons) that are capable of tumour cellkilling. Intraperitoneal GM-CSF in patientswith ovarian cancer is currently being assessedand similar consideration could be given tomalignant pleural effusions in patients withmesothelioma. Another new approach wouldbe to use inhibitors of haematopoiesis withcancer chemotherapy. Several short (3-5) chainamino acid peptides inhibit cell proliferationand some inhibit normal haematopoiesis. Suchinhibitors when clinically available couldprotect the bone marrow by reducing normalbone marrow cell cycling and thereby thesensitivity of the marrow to the effects ofchemotherapy.The use of recombinant DNA technology

has produced a range of colony stimulatingfactors, which are now available for clinicalstudy. The early results indicate that thesefactors substantially ameliorate some of thetoxic effects ofchemotherapy with conventionaland even accelerated or high dose treatment.Fuller exploration of the factors currentlyavailable and those likely to be available in thenear future will require some carefully de-signed clinical studies. We must hope not onlythat the colony stimulating factors will improvethe safety of cancer chemotherapy but also,because they allow greater flexibility inchemotherapy, that the problems of dose inten-sity can be addressed and survival and otherbenefits examined. It is fitting that the firstclinical study with a colony stimulating factortook place in patients with the most commonmalignancy in Britain. Further studies in theseand other patients are already providing veryvaluable information and should lead toimprovements in the management of patientswith other cancers. The development ofrecombinant haematopoietic factors is a verypotent illustration of the benefit to be obtainedwhen basic science is linked with cancermedicine.

I would like to thank Mrs Eileen Morgan for her assistance withtyping.

1 Steward WP, Thatcher N, Kaye SB. Clinical applications ofmyeloid colony stimulating factors. Cancer Treat Rev1990;17:77-87.

2 Metcalf D. Haemopoietic growth factors-1. Peptideregulatory factors. Lancet 1989;i:825-7.

3 Metcalf D. Haemopoietic growth factors-2. Clinicalapplications. Lancet 1989;i:885-7.

4 Cox R, Musial T, GydeOHB. Reduced erythropoietin levelsas a cause of anaemia in patients with lung cancer. Eur JCancer Clin Oncol 1986;22:511-4.

5 Miller CB, Richard RJ, Piantadosi S et al. Decreasederythropoietin response in patients with the anemia ofcancer. N Engl J Med 1990;322:1689-92.

6 de Campos ES, Radford JA, Steward WP, Hill R, ThatcherN. The effect of rHu erythropoietin (Epo) on levels ofhaemoglobin and number of red cell and platelet trans-fusions during intensive chemotherapy for small cell lungcarcinoma (SCLC) [abstract]. BJ Cancer (BACR/ACPannual meeting) 1991;63:9.

7 Bronchud MH. Scarffe JH, Thatcher N et al. Phase I/IIstudy of recombinant human granulocyte colony-stimulating factor in patients receiving intensive chemo-therapy for small cell lung cancer. Br J Cancer1987;56:809-13.

8 Gabrilove JL, Jakubowski A, Scher H, et al Effect ofgranulocyte colony-stimulating factor on neutropenia andassociated morbidity due to chemotherapy for transitionalcell carcinoma of the urothelium. N Engi J Med1988;318:1414-22.

9 Crawford J, Ozer H, Stoller R, et al. Reduction by granu-locyte colony-stimulating factor of fever and neutropeniainduced by chemotherapy in patients with small-cell lungcancer. N Engl J Med 1991;325:164-70.

10 Amgen Roche Ltd. Neupogen product monograph. Maccles-field, Cheshire: Gardiner-Caldwell Communications Ltd,1991:36-42.

11 Morstyn G, Campbell L, Lieschke G, et al. Treatment ofchemotherapy-induced neutropenia by subcutaneouslyadministered granulocyte colony-stimulating factor withoptimization of dose and duration of therapy. J Clin Oncol1989;7: 1554-62.

12 Morstyn G, Campbell L, Souza LM, et al. Effect ofgranulocyte colony stimulating factor on neutropeniainduced by cytotoxic chemotherapy. Lancet 1988;i:667-72.

13 Eguchi K, Sasaki S, Tamura T, et al. Dose escalation studyof recombinant human granulocyte-colony-stimulatingfactor (KRN8601) in patients with advanced malignancy.Cancer Res 1989;49:5221-4.

14 Bronchud MH, Potter MR, Morgenstem G, et al. In vitroand in vivo analysis of the effects of recombinant humangranulocyte colony-stimulating factor in patients. Br JCancer 1988;58:64-9.

15 Groopman JE. Status of colony stimulating factors in cancerand AIDS. Semin Oncol 1990;17(suppl 1):31-7.

16 Steward WP, Scarffe JH, Austin R, et al. Recombinanthuman granulocyte macrophage colony stimulating factor(rhGM-CSF) given as daily short infusions-a phase Idose-toxicity study. Br J Cancer 1989;59:142-5.

17 Lieschke GJ, Maher D, Cohen J, et al. Effects of bacteriallysynthesised recombinant human granulocyte-macrophagecolony-stimulating factor in patients with advancedmalignancy. Ann Intern Med 1989;110:357-64.

18 Appelbaum FR. The clinical use of hematopoietic growthfactors. Semin Hematol 1989;26(suppl 3):7-14.

19 Antman KS, Griffin JD, Elias A, et al. Effect of recombinanthuman granulocyte-macrophage colony-stimulating fac-tor on chemotherapy-induced myelosuppression. N EnglJMed 1988;319:593-8.

20 Anderson H, Gurney H, ThatcherN, Sindwell R, ScarffeJH,Weiner J. Recombinant human GM-CSF in small celllung cancer: A phase I/II study. Recent Results in cancerResearch 1990;121:155-61.

21 Drings P, Fischer JR. Biology and clinical use of GM-CSFin lung cancer. Lung 1990;Suppl:1059-68.

22 Miyagawa K, Chiba S, Shibuya K, et al. Frequent expres-sion of receptors for granulocyte-macrophage colony-stimulating factor on human nonhematopoietic tumor celllines. J Cell Phys 1990;143:483-7.

23 Vellenga E, Biesma B, Meyer C, Wagteveld L, Esselink M,de Vries EGE. The effects of five hematopoietic growthfactors on human small cell lung carcinoma cell lines:interleukin 3 enhances the proliferation in one of theeleven cell lines. Cancer Res 1991;51:73-6.

24 Avalos BR, Gasson JC, Hedvat C, et al. Human granulocytecolony-stimulating factor: biologic activities and receptorcharacterization on hematopoietic cells and small cell lungcancer cell lines. Blood 1990;75:851-7.

25 Dodwell DJ, Gurney H, Thatcher N. Dose intensity incancer chemotherapy. Br J Cancer 1990;61:789-94.

26 Bronchud MH, Howell A, Crowther D, Hopwood P, SouzaL, Dexter TM. The use of granulocyte colony-stimulatingfactor to increase the intensity of treatment with doxo-rubicin in patients with advanced breast and ovariancancer. Br I Cancer 1989;60:121-5.

27 Thatcher N, Lind M, Stout R, et al. Carboplatin, ifosfamideand etoposide with mid-course vincristine and thoracicradiotherapy for "limited" stage small cell carcinoma ofthe bronchus. BrJ Cancer 1989;60:98-101.

28 Prendville J, Radford J, Thatcher N, et al. Intensive therapyfor small cell lung cancer using carboplatin alternatingwith cis-platin, ifosfamide, etoposide, mid-cycle vincris-tine and radiotherapy. J Clin Oncol 1991;9:1446-52.

125



j.com/


ebruary 1992. Dow

nloaded from


Thatcher

29 Ardizzoni A, Sertoli MR, Corcione A, et al. Acceleratedchemotherapy with or without GM-CSF for small celllung cancer: a non-randomised pilot study. Eur J Cancer1990;26:937-41.

30 Neidhart J, Mangalik A, Kohler W, et al. Granulocytecolony-stimulating factor stimulates recovery of granu-locytes in patients receiving dose-intensive chemotherapywithout bone marrow transplantation. J Clin Oncol1989;7:1685-92.

31 Sheridan WP, Morstyn G, Wolf M, et al. Granulocytecolony-stimulating factor and neutrophil recovery afterhigh-dose chemotherapy and autologous bone marrow.transplantation. Lancet 1989;ii:891-5.

32 Brandt SJ, Peters WP, Atwater SK, et al. Effect ofrecombin-ant granulocyte macrophage colony stimulating factor onhematopoietic reconstitution after high dose chemo-therapy and autologous bone marrow transplantation. NEnglJ Med 1988;318:869-76.

33 Griffin JD. Hemopoietins in oncology: factoring outmyelosuppression. J Clin Oncol 1989;7:151-5.

34 Ravagnani F, Siena S, Bregni M, Sciorelli G, Gianni AM,Pellegris G. Large-scale collection of circulatinghaematopoietic progenitors in cancer patients treated withhigh-dose cyclophosphamide and recombinant humanGM-CSF. Eur J Cancer 1990;26:562-4.

35 Gianni AM, Siena S, Bregni M, et al. Granulocyte-macro-phage colony-stimulating factor to harvest circulatinghaemopoietic stem cells for autotransplantation. Lancet1989;ii:580-5.

36 Gianni AM, Bregni M, Siena S, et al. Rapid and completehemopoietic reconstruction following combined trans-plantation of autologous blood and bone marrow cells. Achanging role for high dose chemo-radiotherapy? HematolOncol 1989;7:139-48.

37 Henon PR, Butturini A, Gale RP. Blood-derivedhaematopoietic cell transplants: blood to blood? Lancet1991;337:961-3.

38 Shea TC, Mason JR, Storniolo AM, et al. Beneficial effectfrom sequential harvesting and reinfusing of PBSC inconjunction with rHuGM-CSF and high dose carbo-platin. Blood 1990;76(suppl 1):165.

39 Glover DJ, Glick JH, Weiler C, Hurowitz S, KligermanMM. WR-2721 protects against the haematological toxi-city of cyclophosphamide: a controlled Phase II trial. JClin Oncol 1986;4:584-8.

40 Glover DJ, Glick JH, Weiler C, Fox K, Guerry D. WR-2721and high-dose cisplatin: an active combination in thetreatment of metastatic melanoma. J Clin Oncol1987;5:574-8.

126



j.com/


ebruary 1992. Dow

nloaded from


haematopoietic growthfactors andlung - thoraxshort bolus intravenous injections, subcutan-_ eous...

Documents