`
`COMPARISON OF TREATMENTS FOR ISLET-CELL CARCINOMA - MOERTEL ET AL.
`
`519
`
`STREPTOZOCIN-DOXORUBICIN, STREPTOZOCIN-FLUOROURACIL, OR
`CHLOROZOTOCIN IN THE TREATMENT OF ADVANCED ISLET·CELL CARCINOMA
`
`CHARLES G. MOERTEL, M.D., MYRTO LEFKOPOULO, Sc.D., STUART LIPSITZ, Sc.D., RICHARD G. HAHN, M.D.,
`AND DAVID KLAASSEN, M.D.
`
`Abstract Background. The combination of streptozo(cid:173)
`cin and fluorouracil has become the standard therapy
`for advanced islet-cell carcinoma. However, doxorubicin
`has also been shown to be active against this type
`of tumor, as has chlorozotocin, a drug that is struc(cid:173)
`turally similar to streptozocin but less frequently causes
`vomiting.
`In this multicenter trial, we randomly as(cid:173)
`Methods.
`signed 105 patients with advanced islet-cell carcinoma to
`receive one of three treatment regimens: streptozocin plus
`fluorouracil, streptozocin plus doxorubicin, or chlorozoto(cid:173)
`cin alone. The 31 patients in whom the disease did not
`respond to treatment were crossed over to chlorozotocin
`alone or to one of the combination regimens.
`Results. Streptozocin plus doxorubicin was superior
`to streptozocin plus fluorouracil in terms of the rate of tu(cid:173)
`mor regression, measured objectively (69 percent vs. 45
`percent, P = 0.05), and the length of time to tumor pro(cid:173)
`gression (median, 20 vs. 6.9 months; P = 0.001). Strepto(cid:173)
`zocin plus doxorubicin also had a significant advantage in
`terms of survival (median, 2.2 vs. 1.4 years; P = 0.004)
`
`that was accentuated when we considered long-term sur(cid:173)
`vival (>2 years). Chlorozotocin alone produced a 30 per(cid:173)
`cent regression rate, with the length of time to tumor
`progression and the survival time equivalent to those
`observed with streptozocin plus fluorouracil. Crossover
`therapy after the failure of either chlorozotocin alone or
`one of the combination regimens produced an overall re(cid:173)
`sponse rate of only 17 percent, and the responses were
`transient. Toxic reactions to all regimens included vomit(cid:173)
`ing, which was least severe with chlorozotocin; hemato(cid:173)
`logic depression; and, with long-term therapy, renal insuf(cid:173)
`ficiency.
`Conclusions. The combination of streptozocin and
`doxorubicin is superior to the current standard regimen of
`streptozocin plus fluorouracil in the treatment of advanced
`islet-cell carcinoma. Chlorozotocin alone is similar in effi(cid:173)
`cacy to streptozocin plus fluorouracil, but it produces few(cid:173)
`er gastrointestinal side effects than the regimens contain(cid:173)
`ing streptozocin. It therefore merits study as a constituent
`of combination drug regimens. (N Engl J Med 1992;
`326:519-23. )
`
`ADVANCED carcinoma of the islet cells of the pan-
`1"\.. creas, although rare, presents a kaleidoscope of
`clinical challenges. In addition to the usual problems
`associated with primary and metastatic tumor bulk,
`patients with islet-cell carcinoma may have evidence
`of a variety of hormonal excesses -
`sometimes sub(cid:173)
`clinical, sometimes life-threatening. A special feature
`of islet-cell carcinomas, which should be weighed in
`any treatment decision, is the frequently indolent pro(cid:173)
`gression of disease even after metastasis has occurred.
`Treatment options should always include judicious
`observation. The choice in cases of advanced disease
`is strongly influenced by the distribution and bulk
`of tumor, the aggressiveness of the disease, and the
`
`From the Mayo Clinic, Rochester, Minn. (C.G.M., R.G.H.); the Dana-Farber
`Cancer Institute, Boston (M.L., S.L.); and the Cancer Control Agency of British
`Columbia, Vancouver, Canada (O.K.). Address reprint requests to Dr. Moertel at
`the Mayo Clinic, Rochester, MN 55905.
`Performed for the Eastern Cooperative Oncology Group (D.C. Tormey, chair).
`Other participating institutions include Albany Medical College, Albany, N.Y.;
`Albert Einstein College of Medicine, Bronx, N.Y.; University of California(cid:173)
`Irvine, Orange; Case Western Reserve University, Cleveland; Chicago Medical
`School, Chicago; Fox Chase Cancer Center, Philadelphia; Hahnemann Medical
`College, Philadelphia; Johns Hopkins Oncology Center, Baltimore; Medical Col(cid:173)
`lege of Ohio, Toledo; University of Minnesota, Minneapolis; Mount Sinai Medi(cid:173)
`cal Center, New York; New York University Medical Center, New York; Newark
`Beth Israel Medical Center, Newark, N.J.; Northwestern University Medical
`Center, Chicago; Our Lady of Lourdes Hospital, Binghamton, N. Y.; Hospital of
`the UniverSity of Pennsylvania, Philadelphia; University of Pittsburgh, Pitts(cid:173)
`burgh; University of Rochester Cancer Center, Rochester, N.Y.; Roswell Park
`Memorial Institute, Buffalo, N.Y.; Rush-Presbyterian-St. Luke's Medical Cen(cid:173)
`ter, Chicago; University of Pretoria, Pretoria, South Africa; the Swiss Group,
`Bern, Switzerland; Tufts-New England Medical Center Hospital, Boston; Nat(cid:173)
`alie Warren Bryant Cancer Center, Tulsa, Okla.; and the University of Wisconsin
`Clinical Cancer Center, Madison.
`Supported in part by grants (CA 13650, CA 23318, CA 15947, CA 06594, CA
`14958, CA 14144, CA 18281, CA 13611, CA 16116, CA 20365, CA 17152,
`CA 16395, CA 17145, CA 15489, CA 11083, CA 12296, CA 25988, CA 21692,
`CA 07190, CA 21076, and CA 21115) from the National Cancer Institute.
`
`nature and severity of the associated endocrine syn(cid:173)
`dromes. The appropriate treatment may include sur(cid:173)
`gical reduction of tumor bulk, hepatic-artery occlu(cid:173)
`sion for disease that is predominantly in the liver,
`specific end-organ blockade for endocrine syndromes
`(e.g., omeprazole for gastrinoma), suppression of hor(cid:173)
`mone production (e.g., octreoride for the vasoactive
`intestinal peptide syndrome), and usually last in line,
`systemic cytotoxic therapy.
`Streptozocin has been marketed specifically for the
`treatment of islet-cell carcinoma; it induces objective(cid:173)
`ly measurable tumor regression in about one third of
`patients. I In an earlier study, our group found that a
`combination of fluorouracil with streptozocin pro(cid:173)
`duced a significant improvement in response rate and
`a trend toward improved survival when compared
`with the use of streptozocin alone. 2 We also found that
`doxorubicin was active in patients in whom previous
`chemotherapy had failed. 3 These findings led to the
`consideration of a combination of doxorubicin and
`streptozocin as therapy for this type of cancer. A pilot
`study found that these agents could be combined safe(cid:173)
`ly at essentially full doses of each (unpublished data).
`Chlorozotocin is a new drug that is structurally very
`similar to streptozocin; both contain a nitrosourea
`moiety. Its toxic effects include hematologic depres(cid:173)
`sion, but with less nausea and vomiting than are asso(cid:173)
`ciated with streptozocin. 4,5 It also has the practical
`advantage of being administered in a single intrave(cid:173)
`nous dose during each course of treatment.
`The present study compared a combination of dox(cid:173)
`orubicin and streptozocin with the more commonly
`used combination of fluorouracil and streptozocin in
`metastatic islet-cell cancer. We also compared the new
`
`The New England Journal of Medicine
`
`Downloaded from nejm.org on November 9, 2015. For personal use only. No other uses without permission.
`
` Copyright © 1992 Massachusetts Medical Society. All rights reserved.
`
`NOVARTIS EXHIBIT 2004
`Roxane v. Novartis, IPR 2016-01461
`Page 1 of 5
`
`
`
`520
`
`THE NEW ENCLAND JOURNAL OF MEDICINE
`
`Feb. 20, 1992
`
`agent chlorozotocin with the two streptozocin combi(cid:173)
`nations.
`
`METHODS
`
`The patients enrolled in this trial were cared for at a number of
`different centers; the study was conducted by the Eastern Coopera(cid:173)
`tive Oncology Croup (ECOC). All those enrolled in the study had
`histologic proof of unresectable or metastatic islet-cell carcinoma.
`Tissue diagnosis was confirmed by pathological review. It was re(cid:173)
`quired that each patient have a measurable indicator of response to
`therapy, which could include evidence of tumor on physical exami(cid:173)
`nation or chest films or well-defined metastatic lesions in the liver on
`radioisotope or CT scanning, provided these lesions measured at
`least 5 cm at their widest point. Malignant hepatomegaly could be
`used as an indicator if there was a clearly palpable liver edge at least
`5 cm below the xiphoid process or the costal margins during quiet
`respiration. For patients without measurable tumor, laboratory as(cid:173)
`says demonstrating excessive hormone production could be accept(cid:173)
`ed as the only indicators of response. This was the case for only
`seven patients. The criteria for exclusion were an ECOC perform(cid:173)
`ance score of 4 (indicating total disability), severe nutritional im(cid:173)
`pairment, recent major su~gery (within three weeks), previous ther(cid:173)
`apy with any of the study agents, any chemotherapy or radiation
`therapy within the previous month, active infection, a leukocyte
`count <4X)09 per liter or a platelet count < 150x)09 per liter,
`active heart disease, a serum creatinine level> 132.6 mmol per liter
`(1.5 mg per deciliter) or a blood urea nitrogen level> I O. 7 mmol per
`liter (30 mg per deciliter), or any elevation of serum bilirubin.
`Patients were also excluded if they had any other concurrent or
`recent malignant disease except cutaneous epitheliomas or cervical
`carcinoma in situ.
`
`Randomization Procedures
`
`Patients were stratified according to ECOC performance score
`and according to whether their indicator of response was measur(cid:173)
`able tumor or endocrine abnormalities on laboratory assays. They
`were then randomly assigned to therapy with chlorozotocin alone,
`fluorouracil plus streptozocin, or doxorubicin plus streptozocin.
`If the initially assigned treatment failed, the patients treated
`with chlorozotocin alone were randomly assigned to receive either
`fluorouracil plus streptozocin or doxorubicin plus streptozocin.
`Those originally assigned to a streptozocin regimen received chloro(cid:173)
`zotocin.
`
`reduced by at least 50 percent. For malignant hepatomegaly, regres(cid:173)
`sion was defined as a reduction of at least 30 percent in the sum of
`the measurements below the xiphoid process and the costal mar(cid:173)
`gins. For hormonal assays, it was defined as a reduction of at least
`50 percent in pretreatment abnormalities. If both tumor size and
`endocrine abnormalities were used as indicators, the response to
`treatment was determined on the basis of measurable tumor. Dis(cid:173)
`ease progression was defined as an increase of at least 25 percent in
`tumor measurements or the detection of new areas of malignant
`disease.
`
`Statistical Analysis
`
`Analysis of disease progression and survival was based on a pro(cid:173)
`portional-hazards multivariate model, used to evaluate the rela(cid:173)
`tion of differences among treatment groups to characteristics of
`the patients and the disease.6 Analyses of differences in rates of re(cid:173)
`gression were modeled by multivariate linear logistic regression. 7
`Unadjusted medians for the lengths of time to various events
`were estimated from Kaplan-Meier life-table curves," and differ(cid:173)
`ences in time distributions that were not adjusted for patients'
`characteristics were evaluated with the log-rank test." For the com(cid:173)
`parison of rates of toxicity and response to treatment, we used
`the Kruskal-Wallis exact test for ordered contingency tables."'·11
`For comparisons of patients' characteristics we used the RxC
`contingency-table exact test, which is similar to Fisher's exact
`test. 12 Only P values of less than 0.10 are reported. All tests were
`two-sided.
`
`RESULTS
`Between November 1978 and June 1985, 125 pa(cid:173)
`tients were enrolled in this study. Eighteen patients
`were subsequently found to be ineligible (eight ran(cid:173)
`domly assigned to receive chlorozotocin, seven to re(cid:173)
`ceive fluorouracil plus streptozocin, and three to re(cid:173)
`ceive doxorubicin plus streptozocin). An additional
`two patients withdrew from the study. Up-to-date fol(cid:173)
`low-up information was available'On all the remaining
`105 patients, except I who was lost to follow-up after
`eight years. As Table I shows, the characteristics of
`the patients who could be evaluated were reasonably
`well balanced among the treatment groups.
`
`Treatment
`
`Results of Treatment
`
`Chlorozotocin was given in a single intravenous injection (150 mg
`per square meter of body-surface area), repeated every seven weeks.
`For the combination regimens, streptozocin was given by intrave(cid:173)
`nous injection at a dose of 500 mg per square meter per day for five
`consecutive days, repeated every six weeks. Fluorouracil was given
`by intravenous injection at a dose of 400 mg per square meter per
`day for five days concurrently with streptozocin. Doxorubicin was
`given along with streptozocin by intravenous injection at a dose of
`50 mg per square meter on days I and 22 of each six-week treatment
`cycle, with a maximal total dose of 500 mg per square meter. Leuko(cid:173)
`cyte and platelet counts were obtained weekly; serum creatinine
`measurements and urinalyses were performed before each cycle of
`therapy. Dosages were reduced if patients had severe nausea or
`vomiting, stomatitis, diarrhea, leukopenia, or thrombocytopenia. If
`the creatinine level became elevated or persistent proteinuria devel(cid:173)
`oped, the dose of streptozocin or chlorozotocin was reduced. If
`these abnormalities persisted, treatment with these agents was dis(cid:173)
`continued. Therapy was continued until disease progression was
`noted.
`.
`
`Evaluation of Response
`
`Patients treated with the drug combinations were reevaluated
`every six weeks, and those treated with chlorozotocin were reevalu(cid:173)
`ated every seven weeks. Measurable tumor masses were considered
`to have regressed if the product of perpendicular diameters was
`
`Among the 105 eligible patients who underwent
`treatment, 3 did not have serial measurements ade(cid:173)
`quate to determine their response. Results for the re(cid:173)
`maining 102 are shown in Figure I. The overall differ(cid:173)
`ences in rates of regression among treatment groups
`are highly significant (P = 0.005). The doxorubicin(cid:173)
`streptozocin group had significantly higher rates of
`regression than either the fluorouracil-streptozocin
`group (P = 0.05) or the chlorozotocin group (P =
`0.002). The median durations of regression, measured
`from the start of therapy to the last observation
`of regression, were 17 months for chlorozotocin, 14
`months for fluorouracil plus streptozocin, and 18
`months for doxorubicin plus streptozocin. By an alter(cid:173)
`native measurement, from the first observation of re(cid:173)
`gression to relapse, the median durations of regression
`were 21 months, 13 months, and 22 months, respec(cid:173)
`tively. There was a larger number of patients with
`very long regressions (lasting two or more years) in the
`doxorubicin-streptozocin group (11 patients, vs. 4 in
`the chlorozotocin group and 2 in the ftuorouracil-
`
`The New England Journal of Medicine
`
`Downloaded from nejm.org on November 9, 2015. For personal use only. No other uses without permission.
`
` Copyright © 1992 Massachusetts Medical Society. All rights reserved.
`
`NOVARTIS EXHIBIT 2004
`Roxane v. Novartis, IPR 2016-01461
`Page 2 of 5
`
`
`
`Vol. 326 No.8
`
`COMPARISON OF TREATMENTS FOR ISLET-CELL CARCINOMA - MOERTEL ET AL.
`
`521
`
`streptozocin group). Six patients treated with doxoru(cid:173)
`bicin plus streptozocin (17 percent) had regressions
`that persisted for more than four years, the longest of
`which lasted seven years and nine months.
`Table 2 shows the relation of tumor regression to
`patient characteristics. Regression rates were higher
`for middle-aged patients (40 to 60 years of age) , but
`this difference was only of borderline significance.
`Thete was no strong evidence that any specific type of
`endocrine abnormality made a patient either more re(cid:173)
`sponsive or more resistant to chemotherapy; the num(cid:173)
`bers were small, however.
`Figure 2 shows the length of time to tumor pro(cid:173)
`gression. Doxorubicin plus streptozocin had a strong
`
`Table 1. Characteristics of the Study Patients.
`
`CHARACTERISTIC
`
`CHLOROZOTOCIN
`
`FLUOROURACIL +
`STREPTOZOCIN
`
`DoXORUBICIN +
`STkEPTOZOCIN
`
`Sex (M/F)
`Median age -
`yr (range)
`Time from diagnosis
`< 3 mo
`3-<12 mo
`;;' 12 mo
`ECOG performance score
`0-1
`2-3
`Endocrine abnormalityt
`Gastrin
`5-HIAA
`Hypercalcemia
`Insulin
`Corticotropin
`VIP
`Glucagon
`None recorded
`Indicator of response
`Tumor size
`Hormone assay
`Both
`
`13/20
`57 (25-80)
`
`20114
`51 (18- 78)
`
`18/20
`53 (16-75)
`
`24
`3
`6
`
`23
`10
`
`4
`5
`2
`3
`I
`2
`I
`16
`
`18
`2
`13
`
`22
`6
`6
`
`24
`10
`
`II
`2
`2
`I
`4
`I
`I
`19
`
`19
`3
`12
`
`19
`14
`5
`
`27
`II
`
`II
`I
`3
`2
`I
`I
`2
`20
`
`20
`2
`16
`
`*The ECOG score is on a scale from 0 (fully aClive) 1o 4 (Iolally disabled).
`tNine patients had more than one endocrine-funclion abnonnality. 5-HIAA denotes
`5-hydroxyindoleacetic acid, and VIP va'Wactive intestinal peptide.
`
`advantage in these terms over the other regimens
`(P = 0.001 for both comparisons), and this advantage
`was sustained over six years of observation. The
`P values remained unchanged after multivariate Cox
`regression analysis.
`Survival times are compared in Figure 3. The chlor(cid:173)
`ozotocin and fluorouracil-streptozocin groups had
`essentially the same survival times (median survival,
`1.5 and 1.4 years) . The patients who received doxoru(cid:173)
`bicin plus streptozocin, however, had decidedly longer
`survival (median, 2.2 years); this survival advan(cid:173)
`tage was even more striking in the long-term follow(cid:173)
`up data. Table 3 shows survival in relation to a
`number of patient characteristics. Statistically sig(cid:173)
`nificant predictors of survival included the ECOG
`performance score and age. The age effect was quad(cid:173)
`ratic -
`i.e., patients from 40 to 60 years of age
`had longer survival times than older or younger
`patients. Survival differences among the treatment
`groups remained significant after adjustment for im-
`
`rrzI Any regression
`~!Il!ll!ll~ 0
`~
`30 Yo • Complete regression
`
`Chlorozotocin
`(n = 33)
`Fluorouracil + Strep-.~!Il!ll!ll!ll!ll!ll~45%
`tozocin (n = 33)
`Doxorubicin + Strep(cid:173)
`tozocin (n = 36)
`
`o
`
`10 20 30
`
`40
`
`50
`
`60
`
`70 80
`
`Patients with Regression (%)
`
`Figure 1. Rates of Tumor Regression, Measured Objectively, Ac(cid:173)
`cording to Treatment Group.
`
`balances in these prognostic factors by the Cox regres(cid:173)
`sion analysis; that is, the doxorubicin-streptozocin
`regimen continued to be superior to the other two regi(cid:173)
`mens in terms of survival (P<O.OI for both compari(cid:173)
`sons).
`Thirty-one patients were crossed over to alternative
`therapy after their originally assigned treatment failed
`to induce regression. One of these did not have se(cid:173)
`rial measurements adequate to determine response.
`Only 2 of 15 patients who could be evaluated had a
`regression with chlorozotocin after the previous fail(cid:173)
`ure of a streptozocin regimen; the same was true for
`only 3 of 15 patients who received a streptozocin regi(cid:173)
`men as secondary treatment. Regressions during sec(cid:173)
`ondary treatment were transient, persisting for a
`median of less than six months. The median inter(cid:173)
`val to disease progression for all patients receiving
`secondary therapy was 4 months, and the median
`
`Table 2. Response to Therapy According to Pa(cid:173)
`tients' Characteristics. *
`
`CHARACTERISTIC
`
`Sex
`Male
`Female
`Age (yr)
`.;40
`>40-60
`>60
`Time from diagnosis
`<3 mo
`3-<12 mo
`;;.12 mo
`ECOG performance score
`0-1
`2-3
`Endocrine abnormalityt
`Yes
`No
`Specific abnormality
`Gastrin
`5-HIAA
`Hypercalcemia
`Insulin
`Corticotropin
`VIP
`Glucagon
`
`No. WITH
`No. OF
`OBJECTIVE
`PATIENTS REGRESSION (%) P VALUE
`
`50
`52
`
`19
`49
`34
`
`63
`23
`16
`
`72
`30
`
`47
`55
`
`25
`8
`6
`6
`5
`4
`2
`
`26 (52)
`24 (46)
`
`7 (37)
`29 (59)
`14 (41)
`
`30 (48)
`14 (61)
`6 (38)
`
`38 (53)
`12 (40)
`
`23 (49)
`27 (49)
`
`13 (52)
`2 (25)
`5 (83)
`4 (67)
`2 (40)
`2 (50)
`1(50)
`
`NS
`
`0.08
`
`NS
`
`NS
`
`NS
`
`NS
`
`*NS denotes nO( significant. 5-HIAA 5-hydroxyindoleacetic acid. and
`VIP vasoactive intestinal peptide. Objective regression was defined a.~
`regression measured objectively by (he methods described in the text.
`tNine patient .. had more than one endocrine-function abnormality.
`
`The New England Journal of Medicine
`
`Downloaded from nejm.org on November 9, 2015. For personal use only. No other uses without permission.
`
` Copyright © 1992 Massachusetts Medical Society. All rights reserved.
`
`NOVARTIS EXHIBIT 2004
`Roxane v. Novartis, IPR 2016-01461
`Page 3 of 5
`
`
`
`522
`
`THE NEW ENGLAND JOURNAL OF MEDICINE
`
`Feb. 20, 1992
`
`~
`~ 100
`c
`0
`'iii
`UJ
`~
`CJ)
`
`80
`
`60
`
`e a..
`"S
`0
`£ .§
`
`40
`
`20
`
`Q)
`
`UJ
`"E
`~ 0
`a..
`0
`
`\_-
`~ .!,
`
`~-, -, -!.,
`
`Doxorubicin + streptozocin (n = 38)
`Fluorouracil + streptozocin (n = 34)
`Chlorozotocin (n = 33)
`
`-~--\..
`-i ___ <t
`.... --~
`~--------. '-----------,
`'---
`~ .......... .
`4
`
`5
`
`6
`
`7
`
`3
`
`Years after Start of Treatment
`
`Figure 2. Length of Time to Disease Progression, According to
`Treatment Group.
`P<0.001 for the comparison between doxorubicin plus streptozo·
`cin and fluorouracil plus streptozocin; P<0.001 for the compari·
`son between doxorubicin plus streptozocin and chlorozotocin.
`
`both figures less favorable
`survival 12 months -
`than with primary therapy.
`Since the end results of this study depended on the
`evaluation of response, we studied the association of
`endocrine response and tumor-size response. Among
`45 sets of adequate serial measurements of both tumor
`mass and endocrine abnormalities during primary or
`secondary treatment, 21 were recorded as indicating
`objective tumor regressions; 19 of these also showed a
`reduction in abnormal endocrine levels, and in one
`patient the endocrine level was unchanged. The only
`truly discordant results were in a patient who had a
`transient partial regression of malignant hepatomega(cid:173)
`ly (lasting four months) in association with a rising
`serum gastrin level.
`
`Toxic Reactions
`Toxic reactions to the first cycle of therapy are
`shown in Table 4, which is a composite of observa(cid:173)
`tions from both primary and secondary treatment.
`With regard to vomiting, the chlorozotocin regimen
`
`100
`
`~
`~
`Q)
`
`.~ «
`UJ
`"E
`ii a..
`
`Q)
`
`80
`
`60
`
`40
`
`20
`
`0
`0
`
`---- Doxorubicin + streptozocin (n = 38)
`_ •••••• Fluorouracil + streptozocin (n = 34)
`-
`Chlorozotocin (n = 33)
`
`2
`
`3
`
`5
`
`6
`
`7
`
`Years after Start of Treatment
`
`Figure 3. Survival, According to Treatment Group.
`P<O.OO4 for the comparison between doxorubicin plus streptozo(cid:173)
`cin and fluorouracil plus streptozocin; P<O.03 for the comparison
`between doxorubicin plus streptozocin and chlorozotocin.
`
`was distinctly more tolerable. This difference was
`more striking than Table 4 indicates, since nausea and
`vomiting associated with chlorozotocin generally oc(cid:173)
`curred for only a few hours after the single-dose treat(cid:173)
`ment. On the other hand, the two streptozocin regi(cid:173)
`mens were usually associated with nausea or vomiting
`lasting through the entire five days of treatment. Sto(cid:173)
`matitis and diarrhea were essentially nonexistent in
`patients given chlorozotocin alone. Alopecia (data not
`shown) was almost universal with the regimen of dox(cid:173)
`orubicin plus streptozocin. The frequency of hemato(cid:173)
`logic toxicity was roughly comparable among the
`three treatment groups. Thrombocytopenia was more
`frequent with chlorozotocin, and leukopenia was more
`severe with fluorouracil plus streptozocin. As is typical
`for agents containing nitrosourea, there was a definite
`tendency to increased hematologic toxicity with long(cid:173)
`term chlorozotocin therapy. Heart failure developed
`in three patients, possibly as a consequence of doxoru-
`
`Table 3. Length of Survival According to Patients'
`Characteristics.
`
`CHARACTERISTIC
`
`Sex
`Male
`Female
`Age (yr)
`,,;;40
`>40-60
`>60
`Time from diagnosis
`<3 mo
`3-<12 mo
`~12 mo
`ECOG performance score
`0-1
`2-3
`Endocrine abnormality
`Yes
`No
`
`*NS denotes not significant.
`
`No. OF
`PATIENTS
`
`MEDIAN SURVIVAL
`(YR)
`
`P VALVE·
`
`51
`54
`
`20
`50
`35
`
`65
`17
`23
`
`74
`31
`
`49
`56
`
`2.0
`1.5
`
`1.5
`2.1
`1.3
`
`1.6
`1.8
`1.6
`
`2.1
`1.0
`
`2.0
`1.6
`
`NS
`
`0.098
`
`NS
`
`0.002
`
`NS
`
`bicin toxicity; in the case of one of these patients, the
`maximal dose of 500 mg per square meter specified in
`the protocol was exceeded. There was one treatment(cid:173)
`related death, of a patient treated with fluorouracil
`plus streptozocin who had severe leukopenia compli(cid:173)
`cated by sepsis.
`Nephrotoxicity was in most cases limited to a mild
`increase in serum creatinine. Nine patients had severe
`chronic renal insufficiency, however. Seven of them
`had azotemia requiring dialysis; two, both treated
`with streptozocin regimens, had severe Fanconi's syn(cid:173)
`drome. Since in many patients in this study the disease
`did not respond to treatment and they therefore had
`only short exposures to chlorozotocin or streptozocin,
`the figures in Table 4 undoubtedly underestimate the
`potential of these agents for causing chronic renal
`damage. Among the patients who received chlorozoto(cid:173)
`cin, there was a clear relation between the total dose
`and chronic nephrotoxicity. No patient who received a
`total dose of 750 mg of chlorozotocin per square meter
`
`The New England Journal of Medicine
`
`Downloaded from nejm.org on November 9, 2015. For personal use only. No other uses without permission.
`
` Copyright © 1992 Massachusetts Medical Society. All rights reserved.
`
`NOVARTIS EXHIBIT 2004
`Roxane v. Novartis, IPR 2016-01461
`Page 4 of 5
`
`
`
`Vol. 326 No.8
`
`COMPARISON OF TREATMENTS FOR ISLET-CELL CARCINOMA - MOERTEL ET AL.
`
`523
`
`Table 4. Toxic Reactions According to Treatment Group.
`
`ToxIc REAcrION*
`
`CHLOROZOTOCIN
`(N = 51)
`
`FLUOROURACIL + DoXORU81C1N +
`STREPTOZOCIN
`STREPTOZOCIN
`(N = 42)
`(N = 44)
`
`Vomiting
`Any
`Severe
`Stomatitis
`Any
`Severe
`Dianitea
`Any
`Severe
`Leukopeniat
`<4 x 109 cells/liter
`<2 x 109 cellslliter
`Thrombocytopeniat
`< IOOx 109 cells/liter
`<50 x 109 cells/liter
`Creatinine elevation:!:
`>0.03 mgldl
`>1.0 mgldl
`Chronic renal insufficiency
`
`43
`2
`
`0
`0
`
`6
`0
`
`53
`14
`
`22
`6
`
`32
`15
`7
`
`percent
`
`81
`41
`
`19
`5
`
`33
`2
`
`56
`25
`
`8
`6
`
`29
`7
`7
`
`80
`20
`
`5
`0
`
`5
`0
`
`57
`5
`
`0
`0
`
`44
`2
`4
`
`·For vomiti~g. stomatitis, dianilea, leukopenia, and thrombocytopenia. the results are for
`the fillll coo ... oflhempy only. The values forcrealinine elevalion are for all cou"'s. Only the
`palienlS who had adequate documenlalion of IOxic reaclions are included.
`tOnly patients for whom adequate serial measurement, were available are included.
`tro convert values for creatinine 10 millimoles per liler. multiply by 88.4.
`
`or less had this complication, whereas 20 percent of
`those who received 751 to 1000 mg per square meter
`and 56 percent of those who received more than 1000
`mg per square meter had chronic renal failure.
`
`DISCUSSION
`Advanced islet-cell carcinoma has proved to be sur(cid:173)
`a
`prisingly sensitive to a variety of cytotoxic drugs -
`sensitivity not shared by other neuroendocrine can(cid:173)
`cers.13 Streptozocin seems to have a specificity for this
`neoplasm, and we have observed a progressive in(cid:173)
`crease in therapeutic effectiveness as we have moved
`from streptozocin alone to streptozocin plus fluoroura(cid:173)
`cil and then to streptozocin plus doxorubicin. This
`effectiveness seems to be substantial, resulting in pal(cid:173)
`liation of endocrine syndromes and, most important,
`improved survival. The duration of tumor regression
`and of improvement in endocrine syndromes was in
`the main relatively long, with a median of 1.5 years
`and a maximum of almost 8 years.
`Although the combination of streptozocin and dox(cid:173)
`orubicin was the most effective, streptozocin produced
`frequent and often severe nausea and vomiting, which
`compromised both the patients' quality of life and
`their compliance. Chlorozotocin was found to have
`therapeutic activity comparable to that of streptozo(cid:173)
`cin plus fluorouracil, with considerably less emetic ef-
`
`fect and more convenience in administration. This
`agent seems attractive as a possible replacement for
`streptozocin in combination regimens. Other agents
`with apparent activity in islet-cell carcinoma that
`have not yet been evaluated in randomized trials in(cid:173)
`clude dacarbazine and interferon alfa. 14,15
`In the past we have continued chemotherapy indefi(cid:173)
`nitely in patients who have achieved an objective tu(cid:173)
`mor regression, a practice that can be trying for pa(cid:173)
`tients and that sometimes leads them to abandon
`therapy. In addition, such a policy can result in irre(cid:173)
`versible chronic renal failure requiring dialysis. Fu(cid:173)
`ture studies should examine the possibility of con(cid:173)
`tinuing therapy only until the disease has stabilized,
`with maximal tumor regression. Considering the usual
`slow rate of growth of islet-cell tumors, such an ap(cid:173)
`proach may allow patients more freedom from the
`stress of therapy before the malignant disease ad(cid:173)
`vances to the point at which the resumption of therapy
`is justified.
`Clearly, islet-cell carcinoma is responsive to chemo(cid:173)
`therapy, but the chance of long-term survival with
`current regimens remains small. There is strong justi(cid:173)
`fication for continued clinical research into this rare
`disease.
`
`We are indebted to Ms. Helen Ranchuck for her assistance in
`data management.
`
`REFERENCES
`
`I. Broder LE, Carter SK. Pancreatic islet cell carcinoma. II. Results of therapy
`with streptozotocin in 52 patients. Ann Intern Med 1973;79: 108-18.
`2. Moertel eG, Hanley JA, Johnson LA. Streptozocin alone compared with
`streptozocin plus fluorouracil in the trealnlent of advanced islet-cell carcino(cid:173)
`ma. N Engl J Med 1980;303:1189-94.
`3. Moertel CU, Lavin PT, Hahn RG. Phase II trial of doxorubicin therapy for
`advanced islet cell carcinoma. Cancer Treat Rep 1982;66:1567-9.
`4. Hoth D, Woolley P, Green D, Macdonald 1, Schein P. Phase I studies on
`chlorozotocin. Clin Pharmacol Ther 1978;23:712-22.
`5. Kovach JS, Moertel CG, Schutt AJ, et al. A phase I study of chlorozotocin.
`Cancer 1979;43:2189-96.
`6. Cox DR. Regression models and life-tables. 1 R Stat Soc [B]1972;34:187-
`220.
`7. Idem. The analysis of binary data. London: Chapman & Hall, 1970.
`8. Kaplan EL, Meier P. Nonparametric estimation from incomplete observa(cid:173)
`tions. 1 Am Stat Assoc 1958;53:457-81.
`9. Mantel N, Haenszel W. Statistical aspects of the analysis of data from
`retrospective studies of disease. J NatI Cancer Inst 1959;22:719-48.
`10. Agresti A, Wackerly D, Boyett 1M. Exact conditional tests for cross-classi(cid:173)
`fications: approximation of attained significance levels. Psychometrika
`1977;44:75-83.
`II. Mehta CR, Patel NR, Tsiatis AA. Exact significance testing to establish
`treatment equivalence with ordered categorical data. Biometrics 1984;40:
`819-25.
`12. Mehta CR, Patel NR. A network algorithm for performing Fisher's exact
`test in r x c contingency tables. 1 Am Stat Assoc 1983;78:427-34.
`13. Moertel CG. An odyssey in the land of small tumors. 1 Clin Oncol 1987;
`10: 1502-22.
`14. Kessinger A, Foley IF, Lemon HM. Therapy of malignant APUD cell
`tumors: effectiveness of DTIC. Cancer 1983;5 I :790-4.
`15. Erlcksson B, Oberg K, Aim G, et a1. Treatment of malignant endocrine
`tumors with human leukocyte interferon. Lancet 1986;2:1307-9.
`
`The New England Journal of Medicine
`
`Downloaded from nejm.org on November 9, 2015. For personal use only. No other uses without permission.
`
` Copyright © 1992 Massachusetts Medical Society. All rights reserved.
`
`NOVARTIS EXHIBIT 2004
`Roxane v. Novartis, IPR 2016-01461
`Page 5 of 5