immunotoxin, IgG-HD37-dgA, in patients with B-cell ...

From bloodjournal.hematologylibrary.org by guest on July 14, 2011. For personal use only.

1996 88: 1188-1197

A phase I study of bolus versus continuous infusion of the anti-CD19 immunotoxin, IgG-HD37-dgA, in patients with B-cell lymphoma MJ Stone, EA Sausville, JW Fay, D Headlee, RH Collins, WD Figg, M Stetler- Stevenson, V Jain, ES Jaffe, D Solomon, RM Lush, A Senderowicz, V Ghetie, J Schindler, JW Uhr and ES Vitetta

Information about reproducing this article in parts or in its entirety may be found online at: http://bloodjournal.hematologylibrary.org/site/misc/rights.xhtml#repub_requests Information about ordering reprints may be found online at: http://bloodjournal.hematologylibrary.org/site/misc/rights.xhtml#reprints Information about subscriptions and ASH membership may be found online at: http://bloodjournal.hematologylibrary.org/site/subscriptions/index.xhtml

Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036. Copyright 2011 by The American Society of Hematology; all rights reserved.


A Phase I Study of Bolus Versus Continuous Infusion of the Anti-CD19 Immunotoxin, IgG-HD37-dgA, in Patients With B-Cell Lymphoma By Marvin J. Stone, Edward A. Sausville, Joseph W. Fay, Donna Headlee, Robert H. Collins, William D. Figg, Maryalice Stetler-Stevenson, Vinay Jain, Elaine S. Jaffe, Diane Solomon, Richard M. Lush, Adrian Senderowicz, Victor Ghetie, John Schindler, Jonathan W. Uhr, and Ellen S. Vitetta IgG-HD37-SMPT-dgA is a deglycosylated ricin A chain (dgA)containing immunotoxin (IT) prepared by conjugating the monoclonal murine (MoAb) anti-CD19 antibody, HD37, t o dgA using the heterobifunctional hindered disulfide linker, N-succinimidyl-oxycarbonyl-cu-methyl-a-(2-pyridyldithio) toluene (SMPT). In this report, we have used t w o regimens for the administration of IgG-HD37-SMPT-dgA t o patients with non-Hodgkin’s lymphoma (NHL) in two concomitant phase I trials. One trial examined four intermittent bolus infusions administered at 48-hour intervals. The other studied a continuous infusion (Cl) administered over the same 8day period. In the intermittentbolus regimen, the maximum tolerated dose (MTD) was 16 mg/m2/8 d and the dose-limiting toxicity (DLT) consisted of vascular leak syndrome (VLS),aphasia, and evidence of rhabdomyolysis encountered at 24 mg/m2/8 d. Using theCl regimen, the MTD was defined by VLS at 19.2 mg/mz/8 d. At theMTD of both regimens, a novel toxicity, consisting of acrocyanosis with reversible superficial distal digitalskin necrosis in theabsence of overt evidence of systemic vasculitis, occurred in 3 paschedule, there tients. Of 23 evaluable patients on the bolus was 1 persisting complete response (CR; >40 months) and

1partial response (PR). Of 9 evaluable patients on the continuous infusion regimen, there was 1 PR. Pharmacokinetic parameters for the bolusregimen at theMTD showed a mean maximum serum concentration (Cmax) of 1,209 f 430 ng/mL, with a median TI& for allcourses of 18.2 hours (range, 10.0 t o 80.0 hours), a volume of distribution(Vd) of 10.9 L (range, 3.1 t o 34.5 L), and a clearance (CL) of 0.45 L/h (range, 0.13 t o 2.3 L/h). For the Cl regimen at MTD, the mean Cmaxwas 473 ng/mL, with a median TI& for allcourses of 22.8 963 ~t hours (range, 24.1 t o 30.6 hours), a Vd of 9.4 L (range, 4.4 to 19.5L), and a CL of 0.32 L/h (range, 0.12 t o 0.55 L/h). Twenty-five percent of the patients on the bolus infusion regimen and 30% on theCl regimen made antibody against mouse lg (HAMA) and/or ricin A chain antibody (HARA).We conclude that this IT can be administered safely and that both regimens achieve comparable peak serum concentrations at theMTD; these concentrations are similar to those achieved previously using other regimens with IgG-dgA ITS at their respective MTDs. Thus, toxicity is related to theserum level of the IT and does not differ with different targeting MoAbs. 0 1996 b y The American Society of Hematology.

I

part of a chimeric fusion toxin.”.“ Consistent evidence of clinical activity of ITSin lymphoid neoplasms’ suggeststhat this strategy should be pursued. Several clinical studies that have targetedCD19 in patients with B-cell non-Hodgkin’s lymphoma (NHL) have been reported.” Anti-CD19-blocked ricin (bR) IT used the antiB4 monoclonal antibody (MoAb) chemically linked to intact ricin (A + B chain).’‘ The reactivity of this IT with tissues B4 epitope was diminished by chemically lackingthe of the B chain.I7 blockingthenaturalligand-bindingsite Anti-B4-bR induced clinical responses when administered eitheras an intermittent bolus’6 or continuous infusions” and is currently undergoing phase 111 testing after maximal cytoreduction and bone marrow transplantation. The bR IT, which uses theB chain to facilitate the efficient internalization of the ricin A chain,’9 has carbohydrate residues on both the ricin A and B chains that are recognized by liver cells.”’,” Indeed, hepatotoxicity is the dose-limiting toxicity of anti-B4-bR when administered on a bolus schedule.” An alternative strategy used in our studies employs ITS with dgA linked via the hindered disulfide linker, SMPT, to IgG2’.23or viadisulfide exchangetoits Fab’fragment of antibody.24Although chemical manipulations, including purification and deglycosylation of Achain are required for production of these ITS,” dgA-ITs prepared with the antiCD22 MoAb, RFB4, have had notable potency in patients with NHL.”,” Phase I studies of IgG-RFB4-SMPT-dgA and Fab”RFB4-dgA have been completed using an intermittent bolus s ~ h e d u l e ~and ~ . *a~ continuous infusion(CI) phase I study of comparable durationhas also been completedusing IgG-RFB4-SMPT-dgA.2hIn all instances, clinical responses were observed in 20% to 40% of the patients. and the dose-

MMUNOTOXINS (ITS) are targetedtherapeutic agents that use an antibody or antibody fragment totarget either bacterial or plant toxins to neoplastic cells in an effort to increase their therapeutic index and avoid nonspecific toxicity.’,2 Lymphoidneoplasms are particularly attractive targets for ITSbecause the expression of severalmarkersisboth lineage andtissue specific. Hence, adult lymphoidmalignancies frequently express highly restricted epitopes of CD 19,’ CD22:’ CD25,‘ and CD30.7,8The toxins used in previous clinical studies in patients with lymphoid malignancies have included ricin or its A chain’ and the ribosome-inactivating proteins (RIPS), saporin’”or pokeweed-antiviralprotein (PAP).”.” Diphtheria toxin (DAB) has also beenused as From the Baylor University Medical Center and Charles A. Sanmons Cancer Center, Dallas, T X ; the Developmental Therapeutics Program, Clinical Pharmacology Branch, Laboratory of Pathology, National Cancer Institute, Bethesda, MD; and the Cancer Immunobiology Center and Department of Microbiology, The University of Texas Southwestern Medical Center at Dallas, Dallas, TX. Submitted February 25, 1996; accepted April 3, 1996. Extramural portions of the study were supported by National Institutes of Health Grants No. CA28149 and CA41081. Address reprint requests to Ellen S. Vitetta, PhD, Cancer Immunobiology Center and Department of Microbiology, The Universie of Texas Southwestern Medical Center at Dallas, 5323 Harry Hines Blvd, Dallas, 7X 75235-8576. The publication costs ofthis article were defrayed in purr by page charge payment.This article must therefore be hereb.v marked “advertisement” in accordance with 18 U.S.C. section 1734 solely to indicate this fact. 0 1996 by The American Society of Hematology 0006-4971/96/8804-0004$3.00/0

1188

Blood, VOI 88,NO 4 (August 15), 1996: PP 1188-1197


1189

CD19 IMMUNOTOXIN THERAPY

limiting toxicity (DLT) consisted of vascular leak syndrome (VLS). VLS was manifested in its mildest form by urinary sodium retention with hypoalbuminemia and in its most severe form by hypotension, noncardiogenic pulmonary edema with or without transient aphasia, rhabdomyolysis, and renal failure. A major problem associated with antibody-based therapeutic approaches is the phenotypic heterogeneity of tumor cells. Expression of a given antigen can be focal within a tumor or can change during the natural history of a tumor. Expression may also differ at different tumor sites within a patient.28One way to overcome this problem is to use cocktails of ITS directed against different determinants that collectively are represented on the vast majority of tumor cells and on tumor progenitor cells. Indeed, preclinical studies in SCID mice with B-cell lymphoma have shown superior therapeutic effects after treatment with a mixture of the antiCD22 and anti-CD19 ITS as compared with optimal doses of either alone.” Importantly, after chemotherapy, a cocktail of the two ITShas been curative in minimal disease in mice.3o To obtain information that would support the notion of using IgG-HD37-SMPT-dgA in a cocktail with IgG-RFB4SMPT-dgA to treat NHL, we have conducted phase I evaluations of IgG-HD37-SMPT-dgA as a single agent administered bothby intermittent bolus and C1 schedules. As reported here, both schedules perform comparably with respect to toxicity, pharmacokinetics, immunogenicity, and clinical response. MATERIALS AND METHODS

IgG-HD37-SMF’T-dgAwas administered by either intermittent bolus administration (Baylor Hospital, Dallas, TX) or byC1 (National Cancer Institute [NCI], Bethesda, MD) for an exposure period of 8 days. The studies were approved by the institutional review boards of the two institutions, and all patients signed informed consent documents that indicated the experimental nature of the proposed therapy.

IT Preparation and Administration IgG-HD37-SMPT-dgA was prepared as previously described” usingthe murine HD37 MoAb3’ manufactured by Abbott Biotech (Needham, MA), the SMPT crosslinker from Pierce (Rockford, IL), and dgA from Inland Laboratories (Austin, TX).3 The IT was administered by one of the two regimens, using intermittent bolus infusion or CI. At Baylor Hospital, dose levels were 2, 4, 8, 16, and 24 m g h ’ (total dose administered over 8 days). These doses were divided into four equal doses administered every other day by a 4-hour intravenous (IV) infusion. At the NCI, the IT was administered by C1 over 8 days at two dose levels (9.6 mg/mz and 19.2 mg/m’). A test dose of 0.1 mg was administered IV before the initiation of each infusion to detect the presence of preformed antibodies not detected by thehuman-anti-mouse Ig antibody (HAMA) test, which might cause an anaphylactic reaction. This was not considered part of the total dose administered.

Assessment of EfJicacy Tumor areas were determined in square centimeters before treatment and during week 4 of the study by physical examination and by computed tomography scans of’the involved areas. The total tumor area was calculated by adding the products of the largest perpendicular diameters of each lesion. A complete response (CR) wasdefinedas the disappearance of all measurable lesions for 1 month. A partial response (PR) was defined as greater than a 50% reduction in total tumor area for more than 1 month without worsening of any lesion or appearance of a new lesion. Minor responses were recorded for the C1 regimen only and were defined as a clear biologic effect with a total tumor reduction of less than 50% for more than 1 month or more than 50% for less than 1 month. Progressive disease (PD) was defined as an increase in tumor area of more than 50% or the appearance of a clearly defined new lesion.

Assessment of Safety

To be eligible for this study, patients were previously diagnosed with NHL of low or intermediate grade (NCI working formulation) and had failed to respond to at least one conventional combination chemotherapy regimen. Objectively measurable disease was necessary, with CD19 expressed on at least 30%ofthe tumor cells. Previous chemotherapy was completed at least 2 weeks before entry except for maintenance corticosteroids, the dose of which could not be escalated during the study. Patients were ineligible if they had central nervous system (CNS) or pulmonary parenchymal involvement, abnormal liver or renal function, a cardiac ejection fraction less than 35%, severe infection, antimouse antibodies greater than 1 bg/mL, human immunodeficiency virus (HIV) positivity, or an urgent local complication requiring radiation therapy or surgery.

The patient’s physical status was assessed daily during treatment and all adverse experiences recorded and graded from I to IV, using standard toxicity scales and criteria previously reported.26Hematologic parameters were assessed daily andblood chemistries were assessed every other day during treatment. A complete assessment (physical examination, hematology, blood chemistry, and urinalysis) was performed at the end of treatment and again 3 weeks later, at the time of tumor response. VLS is characterized by hypotension, hypoalbumenemia, edema, and weight gain. These were graded as previously reported except for weight gain and edema.26This was graded as follows: grade I, minimal pedal edema; grade 11, weight increase by greater than I O Ib; grade 111, pulmonary edema, responding to support in less than 48 hours and weight gain less than 30 lb; grade IV, weight gain greater than 30 Ib with or without pulmonary compromise requiring ventilator support or nonresponsive to support by 48 hours. The maximum tolerated dose (MTD) wasdefinedasthe dose below which one patient experienced grade 4 toxicity or the dose at whichthree or fewer patients out of six experienced grade 3 toxicity.

Immunophenotyping

Pharmacokinetics

Biopsies were obtained for immunohistochemical or immunocytochemical cytologic analysis of CD19 expression using theHD37 antibody according to standard techniques. Eligibility could also be established by demonstrating the expression of CD19 on the or peripheral monoclonal tumor cell population inbonemarrow blood as assessed by flow cytometry.

Concentrations ofthe IT in patient sera were measured as described previou~ly,~~ except that HD37-dgA and not RFB4-dgA was used as the standard. Pharmacokinetic parameters were determined using both compartmental and noncompartmental methods for both dose regimens and curves were generated by plotting plasma concentrations of IT versus time. Both one- and two-compartment models

Patients


1190

STONE ET AL Table 1. Demographics and Pathology Patients (regimen)

Maldfemale Median age in years (range) Median no. of prior treatments Stage: l and 11/111 and IV Low-grade v intermediate-grade NHL

Continuous

Bolus

614 55 (41-72) 4 (2-9) Oil 0

13/10 49 (20-74) 5 (1-10) 0123 12/11

614

using weighted, nonlinear least squares analysisby ADAPT I1 (BiomedicalSimulationsResource,University of SouthernCalifornia, Los Angeles, CA, copyright 1992) were used. Model selection was determined by Akaike’s Information Criterion” and visualexamination of the differences between measured and fitted concentrations (residuals).Weighting was determined by the inverseof the observation variance. Each infusionof the daily bolus regimen was independently fit. The area under the concentration-time curve (AUC) was calculated using the trapezoidal rule with extrapolation to infinity. ThepercentageoftheAUCextrapolated was lessthan 10% for all patients. Maximum plasma concentration (C,,,,,) was determined values directly from the concentration-time profile. Pharmacokinetic for the bolus infusions are the average of the values after individual infusions. RESULTS

Patients and Demography

Twenty-three patients were treated using the bolus infusion regimen and 10 patients were treated using the C1 regimen. Patient demographics are summarized by regimen in 55 years, respectively. Table 1. The median ages were 49 and All patients had stage I11 or IV NHL.Patients had undergone a median of five previous treatments, respectively. Only 1 patient had undergone a single treatment before entry. Two patients in the lowest dose group of the bolus infusion regimen and 1 in the lower dose group of the C1 regimen were on maintenancecorticosteroids. Thepatients wereevenly divided between low and intermediate gradesof disease (1 8/ 15). There were no significant differences in demographics between patients on the two treatment schedules. Toxicity and Safety

Table 2 lists the maximal toxicity grades experienced by eachpatientaccordingtodose level for boththe C1 and bolus dosing schedules. Table 3 lists all toxicities. CZ. Using the C1 schedule, and with the exception of I nonevaluable patient, dose level I (9.6 mg/m2/8 d) was well tolerated. Thenonevaluablepatienthadextensivedisease that was refractory to both chemotherapy andradiation therapy and involved both retroperitoneal and mesenteric nodes. On day 6 of the infusion, he was noted to have decreased blood pressure, increased creatinine, and abdominal distension. Radiologic studies were compatible with a toxic megacolon. Clostridium difJicile toxin was detected in the stool, which also yielded Aeromonas on culture. He succumbed to refractory hypotension despite pressors, antibiotics, and fluids. This adverse experience can be considered atoxic

death on protocol, but of uncertainrelationship to theIT. Toxicmegacolon isaknowncomplication of C diflici/e infection.Itspresencetherefore confounded the definition escalationafter of DLT. It waselected tocontinuedose accrual of 2 additional patients at that dose level. At dose level I1 ( 19.2 mg/m’/8 d), 3 of the 6 treated patients experienced grade I11 toxicity, thus defining the MTD. The same MTD was observed using IgG-RFB4-SMPT-dgA when administered by CL’” In 2 patients, IgG-HD37-SMPT-dgA produced grade 111 VLS at the MTD. In the third patient who, by day 8 of theinfusion, had obtained dramatic tumor shrinkage with essential resolution of a greater than 12 cm3 subcutaneous suprascapular mass. a novel grade 111 toxicity was observed. In thispatient.a persisting Raynaud’s-like syndrome manifested by distal upper and lower digital coldness, pallor, and acrocyanosis occurred 12 days after completing the first course of the IT. Subsequent administration of the ITatdose level I1 was withheld. Evaluation of the patient at the time of initial presentation with acrocyanosis includednegativestudies for HAMA andHARA. an erythrocyte sedimentation rate (ESR) of 41, a white blood cell (WBC) count of 6,800 cells/pL. 64% PMNS. 0% bands, 6% lymphocytes, 14% monocytes, 15% eosinophils. no evidence of circulating immune complexes, an IgG level of 65 1, an 1gM level of 58. but an 1gA level less than I O mg/dL. Cryoglobulins were undetectable. and the results of circulating complement activation studies showed that CH50 was not depressed, with negative ANA. anti-SSA, anti-SSB, and anti-SM. The syndrome worsened over the 2 weeks after its initial presentation, with persisting cyanosis and pallor refractory to strategies including persistent warming, nifedipine, and topically applied nitrates. Angiography showed smallvessel filling defects that corresponded to the areas of cyanosis. These abnormalities were consistent with either clots or small to medium-size vessel arteritis. Although there was no systemic evidence of vasculitis or compromise of renal or CNS function, an empiric trial of pulse treatment with intravenous methylprednisolone followed by prednisone was administered at the level of 60 mg/d over a period of approximately 5 weeks, but was without obvious benefit. The syndrome resolved steadily over a properiod of 3 to 4 months. When the patient’s disease gressed 6 months into the study, he was retreated with the same IT at 9.6 mg/m2/8 without d, recurrence ofthis toxicity. The syndrome did notrecurwhen the patient was subsequentlytreatedat dose level I1 of IgG-RFB4-SMPT-dgA used as a single agent. Aside from VLS and acrocyanosis, all other adverse events in theevaluable patients onthe C1 regimenwere grade I or I1 and were readily reversible. Although myalgias were noted in two instances in dose level 11, there were no increases in muscle enzymes or hemepositive urine to suggest rhabdomyolysis. Intermittent bolus infusion. In the patients treated on the intermittentbolusinfusion regimen, toxicities observed at the first four dose levels (2, 4, 8, and 16 mg/m’) were mild or moderate (grade I or 11) and related to VLS, fever with tachycardia, or myalgia. Becauseonly 2 patients (both in


1191

CD19 IMMUNOTOXIN THERAPY

Table 2. Correlations of Toxicity, Circulating Tumor Cells, and Pharmacokinetic Parameters Patient No.

Grade

A. Continuous infusion Dose level I: 9.6 mg/m2/192 h 01 02 03 04 Dose level II: 19.2 mg/m2/192 h 05 06t 07 08 09 101 B. Bolus infusion Dose level I: 2 mg/m2 01 02 03 Dose level II: 4 mg/m2 04 05 06 Dose level 111: 8 mg/m2 07 08 09 Dose level I V 16 mg/m2 10 11 12 2011 21 22 2311 Dose level V: 24 mg/m2 13 14 15 16 17 18 19

Maximum Tox

Response (duration)

I

MR (1 mo) SD NE PD

II 111

SD

I II I

PD

Circulating Tumor Cell" -

1,199 25 25 0 61 0

444 66 1 1,575 1,203 543 1,352

23.7 ND 25.0 21.9 30.6 25.6

13.7 ND 4.5 4.7 16.2 4.4

0.43 ND 0.13 0.18 0.39 0.12

58 120 61

9.9 26.7 15.3

23.8 12.9 30.3

1.58 0.41 2.28

NNNA NAJNA

440 744 230

18.7 18.2 20.5

11.3 4.2 13.4

0.55 0.28 0.53

5 1 5/0 37710 NAJNA

86 1 760 256

16.3 16.4 26.0

21.o 10.9 34.2

1.o 0.60 1.08

010

51 1 1,260 1,670 1,007

13.3 20.7 32.5 14.5

26.5 7.2 9.7 8.8

1.76 0.29 0.26 0.45

0/620 NAJNA

I

SD SD SD

-

I I

II II 111

PD PD PD SD

-

-

+ -

~

+ + + ~

+ -

+

111

PD SD

II II

SD SD

+

II IV II

SD

+

SD SD

-

II

111 IV

SD PD

HAMNHARA (ng/mL)

0.55 0.15 ND ND

-

CR (40+ mo) SD SD

Cl 14h)

19.5 5.7 ND ND

SD PR ( 2 mo) SD

I

Vd (L)

24.5 26.7 ND ND

II l I

I

(h)

204 652 222 317

Ill

I I

TI,

930 0 56 0

PR (1.5 mo)* SD SD MR (1.5 mo)

111 I1 II

CHAX (ng/mL)

-

+

+ ~

-

-

-

-

-

1,172 1,636

20.9 14.3

12.1 7.7

0.51 0.43

2,738 2,187 1,859 3,858 1,842 1,515 1,669

37.5 15.1 14.9 17.5 10.0 25.2 80

5.7 6.1 7.2 3.2 6.3 17.1 34.5

0.13 0.32 0.40 0.13 0.50 0.55 0.44

010

o/o NAJNA

o/o NAJNA 010 010

o/o 167,000/21,000 51,333/3,750

o/o

o/o o/o

o/o NAJNA

NNNA 0/1,045

o/o 640,000/240,000 010 2,000,000/200,000 6,000,000/143,000 8,327/1,600 NNNA

Abbreviations: MR. minor response; PD, progressive disease; ND, not determined. * Values other than + or - are cells per milliliter. t Infusion interrupted at 98.5 hours. Patient had exacerbation of a persisting Raynaud's-like phenomenon; steroids were started at 1.5 months and, therefore, the patient was not evaluable for the duration of the response. 0 Infusion interrupted at 151.2 hours. /ITransient (