Bardia 2019

The n e w e ng l a n d j o u r na l of m e dic i n e
Original Article
Sacituzumab Govitecan-hziy in Refractory

Metastatic Triple-Negative Breast Cancer
A. Bardia, I.A. Mayer, L.T. Vahdat, S.M. Tolaney, S.J. Isakoff, J.R. Diamond,
J. O’Shaughnessy, R.L. Moroose, A.D. Santin, V.G. Abramson, N.C. Shah,
H.S. Rugo, D.M. Goldenberg, A.M. Sweidan, R. Iannone, S. Washkowitz,
R.M. Sharkey, W.A. Wegener, and K. Kalinsky
A BS T R AC T
BACKGROUND
Standard chemotherapy is associated with low response rates and short progression- The authors’ full names, academic degrees,
free survival among patients with pretreated metastatic triple-negative breast cancer. and affiliations are listed in the Appendix.
Address reprint requests to Dr. Bardia at
Sacituzumab govitecan-hziy is an antibody–drug conjugate that combines a human- the Massachusetts General Hospital Can-
ized monoclonal antibody, which targets the human trophoblast cell-surface anti- cer Center, Harvard Medical School, Law-
gen 2 (Trop-2), with SN-38, which is conjugated to the antibody by a cleavable linker. rence House 304, 10 N. Grove St., Bos-
ton, MA 02114, or at bardia.aditya@mgh
Sacituzumab govitecan-hziy enables delivery of high concentrations of SN-38 to tumors. .harvard.edu.
METHODS N Engl J Med 2019;380:741-51.
We conducted a phase 1/2 single-group, multicenter trial involving patients with DOI: 10.1056/NEJMoa1814213
advanced epithelial cancers who received sacituzumab govitecan-hziy intravenously Copyright © 2019 Massachusetts Medical Society.
on days 1 and 8 of each 21-day cycle until disease progression or unacceptable

toxic effects. A total of 108 patients received sacituzumab govitecan-hziy at a dose
of 10 mg per kilogram of body weight after receiving at least two previous antican-
cer therapies for metastatic triple-negative breast cancer. The end points included
safety; the objective response rate (according to Response Evaluation Criteria in Solid
Tumors, version 1.1), which was assessed locally; the duration of response; the clini-
cal benefit rate (defined as a complete or partial response or stable disease for at least
6 months); progression-free survival; and overall survival. Post hoc analyses deter-
mined the response rate and duration, which were assessed by blinded independent
central review.
RESULTS
The 108 patients with triple-negative breast cancer had received a median of 3 previ-
ous therapies (range, 2 to 10). Four deaths occurred during treatment; 3 patients
(2.8%) discontinued treatment because of adverse events. Grade 3 or 4 adverse events
(in ≥10% of the patients) included anemia and neutropenia; 10 patients (9.3%) had
febrile neutropenia. The response rate (3 complete and 33 partial responses) was 33.3%
(95% confidence interval [CI], 24.6 to 43.1), and the median duration of response was
7.7 months (95% CI, 4.9 to 10.8); as assessed by independent central review, these
values were 34.3% and 9.1 months, respectively. The clinical benefit rate was 45.4%.
Median progression-free survival was 5.5 months (95% CI, 4.1 to 6.3), and overall
survival was 13.0 months (95% CI, 11.2 to 13.7).
CONCLUSIONS
Sacituzumab govitecan-hziy was associated with durable objective responses in pa-
tients with heavily pretreated metastatic triple-negative breast cancer. Myelotoxic
effects were the main adverse reactions. (Funded by Immunomedics; IMMU-132-01
ClinicalTrials.gov number, NCT01631552.)
n engl j med 380;8 nejm.org February 21, 2019 741

The New England Journal of Medicine
Downloaded from nejm.org on February 20, 2019. For personal use only. No other uses without permission.
Copyright © 2019 Massachusetts Medical Society. All rights reserved.
T
riple-negative breast cancer, which the tumor microenvironment, thereby allowing for
is defined by a lack of tumor-cell expres- the delivery of therapeutic concentrations of the
sion of the estrogen receptor, progesterone drug in bystander cells to which the conjugate has
receptor, and human epidermal growth factor not bound. Sacituzumab-bound tumor cells are
receptor 2 (HER2),1,2 accounts for approximately killed by intracellular uptake of SN-38, and adja-
15% of invasive breast cancers3-5 and is associated cent tumor cells are killed by the extracellular
with aggressive tumor biology and a poor prog- release of SN-38.24
nosis. Triple-negative breast cancer is more com- IMMU-132-01 is a phase 1/2, basket design,
mon in younger women than in older women and open-label, single-group, multicenter trial involv-
in black persons than in persons of other races ing patients with various types of advanced solid
and ethnic groups, and it is often associated with cancers who have received at least one previous
visceral metastases.2,4 therapy for metastatic disease. Preliminary results
Although targeted therapies have benefited pa- for 69 patients with metastatic triple-negative
tients with other subtypes of breast cancer, and breast cancer have been reported.25 In 2016, saci
several targeted therapies for hormone-receptor– tuzumab govitecan-hziy was assigned a “break-
positive and HER2-positive breast cancer have re- through therapy” designation by the Food and
cently been approved for use,6-9 sequential single- Drug Administration (FDA) for the treatment of
agent chemotherapy remains the standard of care patients with metastatic triple-negative breast can-
for patients with metastatic triple-negative breast cer who have received at least two previous thera-
cancer.7,10 Recently, a progression-free survival of pies for metastatic disease, and, accordingly, the
7.2 months was reported with investigational use protocol was amended to require further enroll-
of atezolizumab and nab-paclitaxel combination ment in a more defined population of patients
therapy in previously untreated patients with meta- with metastatic triple-negative breast cancer who
static triple-negative breast cancer.11 However, a had received at least two lines of previous therapy,
majority of patients have disease progression after including previous taxane therapy. Here, we report
receiving first-line therapy, and standard therapeu- updated results for all patients who received saci-
tic options are limited to chemotherapy. Standard tuzumab govitecan-hziy as a third-line or higher
chemotherapy is associated with low response line of therapy for metastatic triple-negative breast
rates (10 to 15%) and short progression-free sur- cancer at the clinically selected dose level of 10 mg
vival (2 to 3 months) among patients with pre- per kilogram of body weight.
treated metastatic triple-negative breast cancer.12-16
Overall survival among patients with this form Me thods
of breast cancer has not changed over the past
20 years6; this highlights the need for advances Trial Design and Oversight
in therapeutic options for these patients. Eligibility criteria that were previously described
Sacituzumab govitecan-hziy (IMMU-132; Im- for patients with metastatic triple-negative breast
munomedics) is an antibody–drug conjugate in cancer25 also apply to the overall trial population
which SN-38 (an active metabolite of irinotecan), (see the Methods section in the Supplementary Ap-
a topoisomerase I inhibitor, is coupled to the hu- pendix, available with the full text of this article at
manized antitrophoblast cell-surface antigen 2 NEJM.org). Sacituzumab govitecan-hziy was ad-
(Trop-2) monoclonal antibody hRS7 IgG1κ through ministered as an intravenous infusion on days 1
the cleavable CL2A linker.17 Trop-2, a transmem- and 8 of 21-day cycles until disease progression
brane calcium signal transducer, is overexpressed or unacceptable adverse events. In case of severe
in many epithelial cancers, and it stimulates can- treatment-related adverse events, dose reductions
cer-cell growth.18,19 Trop-2 is detected in breast of 25% were allowed for the first occurrence and
cancer cells, including those in triple-negative of 50% for the second occurrence, with treatment
breast cancer,19,20 and its expression is reported discontinuation at the third occurrence. Hemato-
in more than 85% of tumors.21,22 On binding to poietic growth factors or blood transfusions were
Trop-2, hRS7 (in free or conjugated form) is inter- allowed at the investigator’s discretion; however,
nalized and delivers SN-38 into the tumor cell.23 prophylactic treatment before the first dose (cy-
In addition, because of the cleavable linker, SN-38 cle 1, day 1) was not allowed. All patients with
is released in tumors both intracellularly and in metastatic triple-negative breast cancer in the
742 n engl j med 380;8 nejm.org February 21, 2019

Sacituzumab Govitecan-hziy for Breast Cancer
efficacy data set received treatment at a starting Table 1. Characteristics of the Patients at Baseline.
dose of 10 mg per kilogram.
The sponsor, Immunomedics, designed the trial Patients
Characteristic (N=108)
and gathered the data. Data analysis was performed
by Veristat and by authors who are employed by Sex — no. (%)
Immunomedics. All the authors assume respon- Female 107 (99.1)
sibility for the accuracy and completeness of the Male 1 (0.9)
data and vouch for the fidelity of the trial to the Median age (range) — yr 55 (31–80)
protocol, which is available at NEJM.org. The first Race or ethnic group — no. (%)*
draft of the manuscript was written by the first White 82 (75.9)
author with the sponsor and a medical writer paid Black 8 (7.4)
by the sponsor. All the authors contributed to the Asian 3 (2.8)
writing, review, and revision of the manuscript. Other or not specified† 15 (13.9)
ECOG performance-status score — no. (%)‡
Efficacy Evaluations 0 31 (28.7)
Staging computed tomography (CT) and magnetic 1 77 (71.3)
resonance imaging (MRI) were performed at base- Previous anticancer regimens — median no. (range) 3 (2–10)
Previous use of taxanes or anthracyclines for metastatic
line and at 8-week intervals from the start of treat- or nonmetastatic disease — no. (%)
ment until disease progression requiring discon- Taxanes 106 (98.1)
tinuation of treatment. Confirmatory CT and MRI Anthracyclines 93 (86.1)
were performed no sooner than 4 weeks after an Previous use of chemotherapy drugs for metastatic disease
initial partial or complete response. Subsequent — no. (%)
imaging was performed at 8-week intervals after Cyclophosphamide 20 (18.5)
the confirmatory imaging. Platinum agents 74 (68.5)
The primary efficacy end point was the objec- Gemcitabine 59 (54.6)
tive response rate. Assessment of response was Fluoropyrimidine agents 56 (51.9)
performed according to Response Evaluation Crite- Eribulin 49 (45.4)
ria in Solid Tumors, version 1.1. Local assessments Vinorelbine 17 (15.7)
were used for treatment decisions and for the pri- Previous use of checkpoint inhibitors — no. (%) 18 (16.7)
mary efficacy analysis. Other efficacy end points Most common sites of disease — no. (%)
were the time to response and the duration of re- Visceral organ§ 83 (76.9)
sponse in patients who had a response, the clinical Lung or pleura 61 (56.5)
benefit rate (defined as a complete or partial Liver 45 (41.7)
response or stable disease for at least 6 months), Other visceral organ: adrenal glands, pancreas, 7 (6.5)
and progression-free and overall survival. and kidneys
Blinded independent central review of staging Nonvisceral site 25 (23.1)
scans was also obtained for the 56 patients (of the * Race or ethnic group was reported by the patients.
108 with metastatic triple-negative breast cancer) † Eleven patients did not indicate or declined to indicate their race or ethnic
who had complete or partial remission, or at least group; four identified as Native American, Dominican, Hispanic, or mixed race.
‡ The Eastern Cooperative Oncology Group (ECOG) performance-status scale
a 20% reduction in the baseline sum of the diam- ranges from 0 to 5, with 0 indicating that the patient is fully active with no re-
eters of the target lesions, according to local site strictions, 1 indicating that the patient is ambulatory and able to carry out
evaluation. This blinded independent review, which work of a light or sedentary nature but restricted in physically strenuous activ-
ity, and higher numbers indicating increasing degrees of disability.
was performed by Intrinsic Imaging, included re- § Visceral organs were defined as solid organs, excluding the brain.
views by two independent radiologists and a third
adjudicating radiologist, if needed.
of the infusion on day 1 of every even-numbered
Safety Evaluations treatment cycle, at the end of treatment, and at
Safety evaluations included assessments of adverse the end of the trial. Adverse events were coded
events and serious adverse events, laboratory safety according to the Medical Dictionary for Regulatory
evaluations, vital signs, physical examination, and Activities, version 20.0, and severity was graded
12-lead electrocardiography (ECG). A 12-lead ECG according to Common Terminology Criteria for
was to be performed at baseline, after completion Adverse Events (CTCAE), version 4.0. For events

Table 2. Summary of Adverse Events in Patients Receiving Sacituzumab Govitecan-hziy.*
Adverse Event Patients (N=108)
Any Grade Grade 3 Grade 4
number of patients with event (percent)

Any adverse event 108 (100) 71 (66) 21 (19)
Gastrointestinal disorders 102 (94) 21 (19) 0
Nausea 72 (67) 7 (6) 0
Diarrhea 67 (62) 9 (8) 0
Vomiting 53 (49) 7 (6) 0
Constipation 37 (34) 1 (1) 0
Abdominal pain† 27 (25) 1 (1) 0
Mucositis‡ 15 (14) 0 0
General disorders and administration-site 82 (76) 10 (9) 0
conditions
Fatigue and asthenia 59 (55) 9 (8) 0
Peripheral edema 17 (16) 0 0
Pyrexia 13 (12) 0 0
Blood and lymphatic system disorders 80 (74) 25 (23) 15 (14)
Neutropenia§ 69 (64) 28 (26) 17 (16)
Anemia 54 (50) 12 (11) 0
Metabolism and nutrition disorders 70 (65) 21 (19) 2 (2)
Decreased appetite 32 (30) 0 0
Hyperglycemia 26 (24) 3 (3) 1 (1)
Hypomagnesemia 23 (21) 1 (1) 0
Hypokalemia 19 (18) 2 (2) 0
Hypophosphatemia 16 (15) 10 (9) 0
Dehydration 14 (13) 4 (4) 0
Skin and subcutaneous tissue disorders 66 (61) 5 (5) 0
Alopecia 39 (36) 0 0
Rash¶ 30 (28) 2 (2) 0
Pruritus 17 (16) 0 0
Dry skin 15 (14) 1 (1) 0
Abnormal values 62 (57) 22 (20) 6 (6)
Decreased white-cell count 23 (21) 9 (8) 3 (3)
Prolonged activated partial thromboplastin time 15 (14) 2 (2) 0
Increased aspartate aminotransferase level 15 (14) 1 (1) 0
Increased alanine aminotransferase level 15 (14) 1 (1) 0
Decreased weight 15 (14) 0 0
Increased blood alkaline phosphatase level 12 (11) 2 (2) 0
Increased blood lactate dehydrogenase level 11 (10) 0 0
Nervous system disorders 59 (55) 4 (4) 0
Headache 23 (21) 1 (1) 0
Dizziness 22 (20) 0 0
Neuropathy‖ 20 (19) 0 0
Dysgeusia 12 (11) 0 0

Table 2. (Continued.)
Adverse Event Patients (N=108)
Any Grade Grade 3 Grade 4
number of patients with event (percent)

Infections and infestations 56 (52) 11 (10) 2 (2)
Respiratory infection** 23 (21) 3 (3) 0
Urinary tract infection 22 (20) 3 (3) 0
Musculoskeletal and connective-tissue disorders 56 (52) 0 0
Back pain 24 (22) 0 0
Arthralgia 17 (16) 0 0
Pain in extremity 11 (10) 0 0
Respiratory, thoracic, and mediastinal disorders 55 (51) 4 (4) 1 (1)
Cough and productive cough 21 (19) 0 0
Dyspnea 20 (19) 2 (2) 1 (1)
Psychiatric disorders 27 (25) 1 (1) 0
Insomnia 15 (14) 0 0
* Shown are the adverse events of any grade (according to Common Terminology Criteria for Adverse Events [CTCAE],
version 4.0) that occurred in at least 10% of the patients. The Medical Dictionary for Regulatory Activities system organ
class and preferred terms are reported whenever possible.
† This category includes abdominal pain, abdominal distention, upper abdominal pain, abdominal discomfort, and ab-
dominal tenderness.
‡ This category includes stomatitis and mucosal inflammation.
§ This category includes neutropenia and decreased neutrophil counts. Febrile neutropenia of all grades was observed
in 10 patients (9%), and grade 3 and grade 4 febrile neutropenia was observed in 7 patients (6%) and 2 patients (2%),
respectively.
¶ This category includes maculopapular rash, generalized rash, dermatitis acneiform, and skin disorder.
‖ This category includes peripheral neuropathy, paresthesia, peripheral sensory neuropathy, and hypoesthesia.
** This category includes upper respiratory tract infection, viral upper respiratory tract infection, lower respiratory tract
infection, pneumonia, influenza, bronchitis, and respiratory syncytial virus infection.
with varying severity, the maximum reported grade with log–log transformation. Subgroup analyses
was used in summaries. The safety evaluation for were used to evaluate the effect of patient factors
the cohort of patients with metastatic triple-neg- and previous cancer treatments.
ative breast cancer was the same as that for the The trial was approved by the institutional
overall trial population. review board at each investigational site before
initiation of the trial and was performed in ac-
Statistical Analysis cordance with the Declaration of Helsinki, the
Details of the statistical analyses have been de- International Council for Harmonisation guide-
scribed previously.25 When the cohort of patients lines for Good Clinical Practice, the FDA Code of
with metastatic triple-negative breast cancer was Federal Regulations, the requirements of national
evaluated to identify those who received sacituz drug and data protection laws, other applicable
umab govitecan-hziy at a dose of 10 mg per kilo- regulatory requirements, and the standard oper-
gram and had received at least two previous ating procedures of Immunomedics. All patients
therapies for metastatic disease, the target group provided written informed consent before en-
included 108 patients. The response rate and the rollment.
exact 95% confidence intervals were calculated
with the use of the Clopper–Pearson method. R e sult s
Progression-free and overall survival and time-
to-event end points were analyzed with the use Patient Population
of Kaplan–Meier methods, with medians and cor- A total of 108 patients with metastatic triple-nega-
responding 95% confidence intervals determined tive breast cancer (median age, 55 years) were en-
according to the Brookmeyer and Crowley method rolled between June 2013 and February 2017. The

patients had been heavily pretreated (median of the investigator not to be drug-related. Transient
3 previous anticancer regimens [range, 2 to 10]). changes in laboratory safety values that occurred
A total of 106 of these patients (98%) had received during treatment included decreases in blood-cell
taxanes and 93 (86%) had received anthracyclines. counts and alterations in biochemical values,
The baseline demographic characteristics of the which generally recovered by the end of treatment.
patients are summarized in Table 1. At the time
of data cutoff (December 1, 2017), the median Efficacy
duration of follow-up among the 108 patients with In Figure 1A, a waterfall plot shows the breadth
metastatic triple-negative breast cancer was 9.7 and depth of responses according to local assess-
months (range, 0.3 to 36.5). Eight of these patients ment in 108 patients with metastatic triple-nega-
were continuing to receive treatment. A total of tive breast cancer. The response rate was 33.3%
100 patients (92.6%) had discontinued treatment, (36 of 108 patients), including complete respons-
and in 86 of these patients (80%), discontinuation es in 3 patients (2.8%). The clinical benefit rate
was because of disease progression. Other reasons (including stable disease for at least 6 months)
for discontinuation are listed in Table S1 in the was 45.4% (49 of 108 patients). In Figure 1B, a
Supplementary Appendix. swimmer plot shows the onset and durability of
response in each of the 36 patients who had an
Safety objective response. In these patients, the median
The 108 patients with metastatic triple-negative time to response was 2.0 months (range, 1.6 to
breast cancer received a mean of 18.7 doses of 13.5), and the median duration of response was
sacituzumab govitecan-hziy (range, 1 to 102), or 7.7 months (95% confidence interval [CI], 4.9 to
9.6 cycles (range, 1 to 51), with a median duration 10.8). The response rate (34.3% [95% CI, 25.4 to
of exposure of 5.1 months (range, 0.03 to 36.1). A 44.0]) and median duration of response (9.1 months
total of 99 patients (92%) received preinfusion [95% CI, 4.6 to 11.3]) according to blinded inde-
medications (acetaminophen, antihistamines, H2 pendent review were similar to those determined
antagonists, glucocorticoids, antiemetics, anxio- by local assessment (Table S5 in the Supplemen-
lytics, and atropine). The most common adverse tary Appendix). According to local assessment, the
events were nausea, diarrhea, fatigue, neutropenia, estimated probability that a patient would have
and anemia, and the most common adverse events a response at 6 months was 59.7%, and the esti-
of grade 3 or higher (>5% incidence) included neu- mated probability that a patient would have a
tropenia, anemia, and a decreased white-cell count, response at 12 months was 27.0% (Fig. S2 in the
as outlined in Table 2. Diarrhea (predominantly Supplementary Appendix). At the database cutoff
grade 1) was a common adverse event (in 62% of date, 6 patients had long-term responses with a
the patients overall); the incidence of CTCAE grade response for more than 12 months (range, 12.7
2 diarrhea was 14%, and the incidence of at least to 30.4). Table 3 summarizes the results according
grade 3 diarrhea was 8%. No peripheral neuropa- to investigator assessment.
thy of grade 3 or higher was reported. Four pa- We evaluated the response to sacituzumab
tients (4%) had adverse events leading to death govitecan-hziy in a variety of patient subgroups
during treatment (details are provided in the (Table S6 in the Supplementary Appendix) and
Results section in the Supplementary Appendix). found no meaningful differences in response rates
Serious adverse events were reported in 35 pa- according to patient age, the onset of metastatic
tients (32%); the most common (>2% incidence) disease, the number of previous therapies, or the
were febrile neutropenia (in 7% of the patients), presence or absence of visceral metastases. The
vomiting (in 6%), nausea (in 4%), diarrhea (in 3%), response rate was 44% (8 of 18) among patients
and dyspnea (in 3%). who had received previous checkpoint inhibitors;
Adverse events leading to interruption of treat- however, these results and those of all the reported
ment occurred in 48 of the 108 patients (44%); the subgroups should be interpreted with caution
most common reason was neutropenia. Three pa- given the small number of patients available,
tients (3%) discontinued treatment because of ad- which led to wide confidence intervals.
verse events; 2 patients discontinued because of At the time of data cutoff, 94 patients (87.0%)
drug-related events, and 1 patient discontinued had disease progression and 77 patients (71.3%)
because of hypertension, which was thought by had died. The median progression-free survival

Figure 1. Response and Survival among 108 Patients A Change in Tumor Size
with Metastatic Triple-Negative Breast Cancer.
Panel A shows a waterfall plot of the best percent 90
Progression Stable Partial Complete
change from baseline in the sum of the diameters of 70 of disease disease response response
the target lesions (longest diameter for non-nodal le-
Change from Baseline (%)

50
sions and short axis for nodal lesions). In 3 patients
30
(2 with stable disease and 1 with progressive disease)
(asterisks), the best percent change was zero. The 10
0 ***
dashed lines at 20% and −30% indicate progressive −10
disease and partial response, respectively, according −30
to Response Evaluation Criteria in Solid Tumors −50
(RECIST), version 1.1. Additional details are provided
−70
in the Results section in the Supplementary Appendix.
Panel B shows a swimmer plot of the objective respons- −90
es (according to RECIST, version 1.1) from the start of −110
treatment to disease progression, as determined by lo-
cal assessment. At the time of the analysis, 6 patients B Patients with Objective Response
had a continuing response. The vertical dashed lines
show the response at 6 months and 12 months, which
are clinically meaningful end points for patients with
metastatic triple-negative breast cancer. Panel C shows
a Kaplan–Meier analysis of progression-free survival
among the 108 patients.
Individual Patients
was 5.5 months (95% CI, 4.1 to 6.3); the estimated

probability of progression-free survival was 41.9%
at 6 months and 15.1% at 12 months (Fig. 1C).
The median overall survival was 13.0 months Complete response
(95% CI, 11.2 to 13.7); the estimated probability Partial response
of survival was 78.5% at 6 months and 51.3% at Ongoing response after data cutoff
12 months (Fig. S3 in the Supplementary Ap- Onset of response
pendix). 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40
To analyze efficacy in relation to the aggres- Treatment Duration (mo)
siveness of the clinical course and to address
concerns regarding relatively indolent tumor bi- C Progression-free Survival
ologic characteristics in the trial population, we 1.0 Median, 5.5 (95% CI, 4.1–6.3)
compared the duration of treatment with sacituz 0.9
Probability of Progression-free
umab govitecan-hziy with that of previous anti- 0.8

cancer treatment in the 108 patients with meta- 0.7
static triple-negative breast cancer for whom data 0.6

Survival
0.5
were available. The median duration of treatment
0.4
with sacituzumab govitecan-hziy (5.1 months) was
0.3
approximately twice that with the previous anti-
0.2
cancer treatment (2.5 months); this highlights
0.1
the clinical activity and lack of cross-resistance 0.0
with this antibody–drug conjugate (Fig. 2). 0 3 6 9 12 15 18 21 24 27 30 33 36
Treatment Duration (mo)
Discussion No. at Risk 108 73 43 22 12 7 5 3 3 1 1 1 0
Among patients with metastatic triple-negative

breast cancer who had received at least two pre- months, the median progression-free survival was
vious therapies for metastatic disease (median, 5.5 months, and the median overall survival was
three) and who received treatment with sacitu- 13.0 months. Efficacy was observed in patients
zumab govitecan-hziy, the response rate was who had received taxanes and anthracyclines, sug-
33.3%, the median duration of response was 7.7 gesting a lack of cross-resistance to previous cyto-

Table 3. Summary of Treatment Efficacy, According to Local Assessment.

treatment with cytotoxic agents in this patient
population,26 were not observed. Hypersensitivity
Patients events of grade 3 or higher that were associated
Variable (N=108)
with infusion of monoclonal antibodies were infre-
Complete response — no. of patients (%) 3 (2.8) quent (in 3 patients [3%]). Four deaths occurred
Partial response — no. of patients (%) 33 (30.6) during treatment (within 30 days after the last dose
Stable disease — no. of patients (%) 40 (37.0) of sacituzumab govitecan-hziy); all deaths were at-
Progressive disease — no. of patients (%) 28 (25.9)
tributed by the investigators to disease progres-
sion, and none were considered to be related to
Not evaluated — no. of patients (%)* 4 (3.7)
sacituzumab govitecan-hziy (see the Results sec-
Objective response rate† tion in the Supplementary Appendix). The safety
No. of patients 36 profile of sacituzumab govitecan-hziy in the 108
% of patients (95% CI) 33.3 (24.6–43.1) patients with metastatic triple-negative breast can-
Clinical benefit rate‡ cer was generally consistent with that in the overall
No. of patients 49 safety population of 420 patients who had a variety
of tumor types (Table S3 in the Supplementary
% of patients (95% CI) 45.4 (35.8–55.2)
Appendix).
Median duration of response (95% CI) — mo 7.7 (4.9–10.8)
The long-term efficacy of the various treat-
* These patients could not be evaluated because of death, transfer to hospice, ment options for patients with metastatic triple-
withdrawal due to clinical progression, or withdrawal due to grade 4 neutrope- negative breast cancer (serial application of single
nia before any radiologic assessment of response. agents) is limited.12-14,16 Poor outcomes seen in pa-
† The objective response rate is the percentage of patients with a complete re-
sponse or partial response. tients with metastatic triple-negative breast cancer,
‡ The clinical benefit rate is the percentage of patients with a complete re- as compared with other breast cancer subtypes,
sponse or partial response or stable disease for at least 6 months. are partly explained by the lack of actionable driver
mutations or established molecular targets, there-
toxic chemotherapy. The duration of treatment by leaving sequential single-agent chemotherapy
with sacituzumab govitecan-hziy was longer than as the main treatment approach.7,10 Sacituzumab
with the immediate previous antitumor therapy govitecan-hziy is an antibody–drug conjugate with
(5.1 months vs. 2.5 months); this provides fur- Trop-2 as the target of recognition; it can deliver
ther evidence of clinical activity in patients with cytotoxic chemotherapy to tumors, including ad-
difficult-to-treat metastatic triple-negative breast jacent cancer cells, in concentrations that are
cancer. Although a subgroup analysis based on higher than those with standard chemotherapy
the patients’ age, the onset of metastatic disease, and may reduce toxic effects in normal tissues
the number of previous therapies, or the presence that do not express the target.24 High expression
or absence of visceral metastases showed no of Trop-2 in triple-negative breast cancer and its
meaningful differences in outcomes, the small association with a poor prognosis suggest that it
number of patients led to wide confidence inter- is a rational therapeutic target in this patient
vals, and thus the homogeneity of clinical out- population.19,21
comes observed in these subgroups is weak and The cytotoxic component of sacituzumab gov-
should be interpreted with caution. itecan-hziy is SN-38, a highly potent topoisomer-
The most relevant adverse events in patients ase I inhibitor and metabolite of irinotecan. The
with metastatic triple-negative breast cancer, as cytotoxic activity of SN-38 delivered through
well as in the larger population of patients with sacituzumab govitecan-hziy is 100 to 1000 times
multiple tumor types who received sacituzumab as high as that of irinotecan.27 In animal models,
govitecan-hziy, included neutropenia and diarrhea, the tumor-to-serum area under the curve ratio for
which were managed with routine supportive care SN-38 was 20 to 40 times as high with sacituz
according to general practice guidelines (i.e., early umab govitecan-hziy as it was with irinotecan,
intervention with granulocyte colony-stimulating whereas concentrations that were 20 to 136 times
factor and early intervention for diarrhea). Few as high as those with irinotecan were delivered
patients discontinued treatment because of adverse into the tumor.24 Clinically, serum concentrations
events. Severe drug-related neuropathy or cardiac of glucuronidated SN-38, the molecular species
adverse events, which may limit the duration of most strongly associated with toxic effects, were

Receipt of last previous therapy Receipt of sacituzumab govitecan-hziy Continuation of treatment
* **
**
30 24 18 12 6 0 6 12 18 24 30 36 42
Months
Figure 2. Duration of Treatment with Sacituzumab Govitecan-hziy and with the Last Previous Therapy in the
108 Patients with Metastatic Triple-Negative Breast Cancer.
Asterisks indicate patients who received therapy for only 1 day. The vertical dashed lines at 6 months and 12 months
show the clinically meaningful end points for patients with metastatic triple-negative breast cancer. When the month
and year were available, a missing start date was imputed as the 15th of the month, and a missing end date was
imputed as the last day of the month.
substantially lower than those of SN-38 with Toxic effects of irinotecan, especially neutrope-
sacituzumab govitecan-hziy and substantially low- nia, have been associated with UGT1A1*28 homo-
er than glucuronidated SN-38 concentrations re- zygosity30,31 and with other variants of UGT1A1 or
ported with irinotecan.28 This may explain the DPYD.32-34 The incidence of neutropenia increased
considerably lower discontinuation rates and clin- numerically with the number of *28 copies, where-
ically relevant lower rates of grade 3 or 4 gastro- as this pattern was not observed for other adverse
intestinal toxic effects in this trial than in trials events of interest such as diarrhea (Table S4 in
of irinotecan. Although toxic effects associated Supplementary Appendix). However, the results
with sacituzumab govitecan-hziy are similar to are based on a retrospective, exploratory analysis,
those of irinotecan, these data suggest that saci- and additional validation is needed before they
tuzumab govitecan-hziy has a better side-effect can be used for clinical decision making related
profile and is less likely to be associated with more to sacituzumab govitecan-hziy.
severe adverse effects.29 The less severe nature of Among the few studies involving pretreated
these adverse events with sacituzumab govitecan- patients with metastatic breast cancer, one of the
hziy was reflected in the low incidences of adverse largest, EMBRACE (Eisai Metastatic Breast Can-
events leading to treatment discontinuation (3%) cer Study Assessing Physician’s Choice Versus
and death during treatment (4%), as well as in the Eribulin), involved 762 patients who had received
absence of grade 3 or 4 neurotoxicity. a median of four previous chemotherapy regimens,

and 19% had triple-negative breast cancer.26 Ad- accelerated approval in other oncology trials,36-38
verse events leading to discontinuation of treat- with only a small portion of indications under the
ment occurred in 13% of the patients receiving accelerated approval program failing to verify
eribulin and 15% receiving treatment of the clinical benefit.39 A confirmatory multicenter, ran-
physician’s choice (any single-agent chemotherapy, domized, phase 3 trial (ASCENT; ClinicalTrials.gov
hormone, or radiotherapy). In the EMBRACE study, number, NCT02574455) is currently recruiting pa-
the incidence of neutropenia was 52% with eribu- tients in North America and Europe to compare
lin (the incidence of grade 4 neutropenia was 24%) sacituzumab govitecan-hziy with the physician’s
and 30% with the physician’s choice (grade 4 neu- choice of four single-agent types of chemotherapy
tropenia, 7%) and the response rate was 12% (capecitabine, gemcitabine, vinorelbine, and er-
with eribulin (duration of response, 4.2 months) ibulin) in patients with metastatic triple-negative
and 5% with the physician’s choice (duration of breast cancer that is refractory or relapsed after
response, 6.7 months). The response rate of 33% at least two previous forms of chemotherapy (in-
and the duration of response of 7.7 months re- cluding a taxane).
ported with sacituzumab govitecan-hziy compare In conclusion, sacituzumab govitecan-hziy
favorably.26 In addition, topoisomerase inhibition (IMMU-132) had efficacy with a 33% response rate
may have advantages over microtubule inhibition in a heavily pretreated population of patients with
in these patients, given that altered DNA repair metastatic triple-negative breast cancer. Diarrhea
pathways are common in triple-negative breast and myelosuppression were the primary adverse
cancer.35 Furthermore, confirmed objective re- events, and discontinuation rates were low.
sponses were noted in patients who had received Editor’s note: After this article went to press, the trial sponsor
was informed by the FDA that the suffix “-hziy” had not yet been
previous programmed death 1–based therapy or formally approved.
programmed death ligand 1–based therapy, sug- Supported by Immunomedics.
gesting a lack of cross-resistance with immune Disclosure forms provided by the authors are available with
the full text of this article at NEJM.org.
checkpoint inhibitors and the potential useful- A data sharing statement provided by the authors is available
ness of combination therapy. with the full text of this article at NEJM.org.
Direct comparison with other chemotherapy We thank the patients and their caregivers, the clinical trial
investigators and their team members who participated in this
approaches was not possible in this trial because trial, Heather Horne and Dr. Pius Maliakal for trial management,
of its noncomparative design. However, we used Dr. Sharon K. Wyhopen for critical review of an earlier version of
the response rate as the primary end point, which the manuscript and participation in discussions, Veristat for as-
sistance with the data analysis, and Dr. Axel Glasmacher of AG
is less subject to bias than progression-free sur- Life Science Consulting and Peloton Advantage for editorial and
vival in a single-group trial and has been used for writing support with an earlier version of the manuscript.
Appendix
The authors’ full names and academic degrees are as follows: Aditya Bardia, M.D., Ingrid A. Mayer, M.D., Linda T. Vahdat, M.D.,
M.B.A., Sara M. Tolaney, M.D., M.P.H., Steven J. Isakoff, M.D., Ph.D., Jennifer R. Diamond, M.D., Joyce O’Shaughnessy, M.D., Re-
becca L. Moroose, M.D., Alessandro D. Santin, M.D., Vandana G. Abramson, M.D., Nikita C. Shah, M.D., Hope S. Rugo, M.D., Da-
vid M. Goldenberg, Sc.D., M.D., Ala M. Sweidan, M.B.A., M.S., Robert Iannone, M.D., Sarah Washkowitz, J.D., Robert M. Sharkey,
Ph.D., William A. Wegener, M.D., Ph.D., and Kevin Kalinsky, M.D.
The authors’ affiliations are as follows: the Massachusetts General Hospital Cancer Center (A.B., S.J.I.) and Dana–Farber Cancer
Institute (S.M.T.), Harvard Medical School, Boston; Vanderbilt–Ingram Cancer Center, Nashville (I.A.M., V.G.A.); Weill Cornell Medical
College (L.T.V.) and New York–Presbyterian–Columbia University Irving Medical Center (K.K.), New York; University of Colorado Can-
cer Center, Aurora (J.R.D.); Texas Oncology, Baylor University Medical Center, US Oncology, Dallas (J.O.); Orlando Health University
of Florida Health Cancer Center, Orlando (R.L.M., N.C.S.); Yale University School of Medicine, New Haven, CT (A.D.S.); University of
California, San Francisco, Helen Diller Family Comprehensive Cancer Center, San Francisco (H.S.R.); Immunomedics, Morris Plains,
NJ (D.M.G., R.I., S.W., R.M.S., W.A.W.); and AIS Consulting, Ann Arbor, MI (A.M.S.).
References
1. Anders CK, Zagar TM, Carey LA. The cancer, including race. Cancer Causes Con- 4. Plasilova ML, Hayse B, Killelea BK,
management of early-stage and metastat- trol 2009;20:1071-82. Horowitz NR, Chagpar AB, Lannin DR.
ic triple-negative breast cancer: a review. 3. DeSantis CE, Fedewa SA, Goding Sau- Features of triple-negative breast cancer:
Hematol Oncol Clin North Am 2013;27: er A, Kramer JL, Smith RA, Jemal A. Breast analysis of 38,813 cases from the Nation-
737-749, viii. cancer statistics, 2015: convergence of in- al Cancer Database. Medicine (Baltimore)
2. Trivers KF, Lund MJ, Porter PL, et al. cidence rates between black and white 2016;95(35):e4614.
The epidemiology of triple-negative breast women. CA Cancer J Clin 2016;66:31-42. 5. Kohler BA, Sherman RL, Howlader N,

et al. Annual report to the nation on the 17. Starodub AN, Ocean AJ, Shah MA, et al. Sacituzumab govitecan (IMMU-132),
status of cancer, 1975-2011, featuring in- al. First-in-human trial of a novel anti- an anti-Trop-2-SN-38 antibody-drug con-
cidence of breast cancer subtypes by race/ Trop-2 antibody-SN-38 conjugate, sacituz jugate for the treatment of diverse epithe-
ethnicity, poverty, and state. J Natl Cancer umab govitecan, for the treatment of di- lial cancers: safety and pharmacokinet-
Inst 2015;107(6):djv048. verse metastatic solid tumors. Clin Cancer ics. Cancer 2017;123:3843-54.
6. Zeichner SB, Terawaki H, Gogineni K. Res 2015;21:3870-8. 29. Camptosar. New York:Pharmacia &
A review of systemic treatment in meta- 18. Ripani E, Sacchetti A, Corda D, Al- Upjohn, 2016 (package insert).
static triple-negative breast cancer. Breast berti S. Human Trop-2 is a tumor-associ- 30. Dean L. Irinotecan therapy and
Cancer (Auckl) 2016;10:25-36. ated calcium signal transducer. Int J Can- UGT1A1 genotype. In:Pratt V, McLeod H,
7. Cardoso F, Costa A, Senkus E, et al. cer 1998;76:671-6. Rubinstein W, et al., eds. Medical genetics
3rd ESO-ESMO international consensus 19. Trerotola M, Cantanelli P, Guerra E, et summaries. Bethesda, MD:National Cen-
guidelines for advanced breast cancer al. Upregulation of Trop-2 quantitatively ter for Biotechnology Information, April
(ABC 3). Breast 2017;31:244-59. stimulates human cancer growth. Onco- 4, 2018 (https://www.ncbi.nlm.nih.gov/
8. Finn RS, Martin M, Rugo HS, et al. gene 2013;32:222-33. books/NBK294473/).
Palbociclib and letrozole in advanced 20. Huang H, Groth J, Sossey-Alaoui K, 31. Kweekel D, Guchelaar HJ, Gelderblom
breast cancer. N Engl J Med 2016; 375: Hawthorn L, Beall S, Geradts J. Aberrant H. Clinical and pharmacogenetic factors
1925-36. expression of novel and previously de- associated with irinotecan toxicity. Can-
9. Verma S, Miles D, Gianni L, et al. scribed cell membrane markers in human cer Treat Rev 2008;34:656-69.
Trastuzumab emtansine for HER2-posi- breast cancer cell lines and tumors. Clin 32. Liu D, Li J, Gao J, Li Y, Yang R, Shen L.
tive advanced breast cancer. N Engl J Med Cancer Res 2005;11:4357-64. Examination of multiple UGT1A and
2012;367:1783-91. 21. Goldenberg DM, Cardillo TM, Govin- DPYD polymorphisms has limited ability
10. NCCN clinical practice guidelines in dan SV, Rossi EA, Sharkey RM. Trop-2 is a to predict the toxicity and efficacy of met-
oncology:breast cancer, version 1.2018. novel target for solid cancer therapy with astatic colorectal cancer treated with iri-
National Comprehensive Cancer Network, sacituzumab govitecan (IMMU-132), an notecan-based chemotherapy: a retrospec-
2018 (https://www.nccn.org/professionals/ antibody-drug conjugate (ADC). Oncotar- tive analysis. BMC Cancer 2017;17:437.
physician_gls/pdf/breast.pdf). get 2015;6:22496-512. 33. Falvella FS, Cheli S, Martinetti A, et
11. Schmid P, Adams S, Rugo HS, et al. 22. Goldenberg DM, Stein R, Sharkey al. DPD and UGT1A1 deficiency in
Atezolizumab and nab-paclitaxel in ad- RM. The emergence of trophoblast cell- colorectal cancer patients receiving triplet
vanced triple-negative breast cancer. surface antigen 2 (TROP-2) as a novel can- chemotherapy with fluoropyrimidines,
N Engl J Med 2018;379:2108-21. cer target. Oncotarget 2018;9:28989-9006. oxaliplatin and irinotecan. Br J Clin Phar-
12. Khosravi-Shahi P, Cabezón-Gutiérrez 23. Cardillo TM, Govindan SV, Sharkey macol 2015;80:581-8.
L, Custodio-Cabello S. Metastatic triple RM, Trisal P, Goldenberg DM. Human- 34. Riera P, Salazar J, Virgili AC, et al.
negative breast cancer: optimizing treat- ized anti-Trop-2 IgG-SN-38 conjugate for Relevance of CYP3A4*20, UGT1A1*37
ment options, new and emerging targeted effective treatment of diverse epithelial and UGT1A1*28 variants in irinotecan-
therapies. Asia Pac J Clin Oncol 2018;14: cancers: preclinical studies in human can- induced severe toxicity. Br J Clin Pharma-
32-9. cer xenograft models and monkeys. Clin col 2018;84:1389-92.
13. Brufsky A, Valero V, Tiangco B, et al. Cancer Res 2011;17:3157-69. 35. Sharma P. Biology and management
Second-line bevacizumab-containing ther- 24. Sharkey RM, McBride WJ, Cardillo of patients with triple-negative breast
apy in patients with triple-negative breast TM, et al. Enhanced delivery of SN-38 to cancer. Oncologist 2016;21:1050-62.
cancer: subgroup analysis of the RIBBON-2 human tumor xenografts with an anti- 36. Shaw AT, Ou S-HI, Bang Y-J, et al.
trial. Breast Cancer Res Treat 2012;133: Trop-2-SN-38 antibody conjugate (sacituz Crizotinib in ROS1-rearranged non–
1067-75. umab govitecan). Clin Cancer Res 2015; small-cell lung cancer. N Engl J Med 2014;
14. Perez EA, Patel T, Moreno-Aspitia A. 21:5131-8. 371:1963-71.
Efficacy of ixabepilone in ER/PR/HER2- 25. Bardia A, Mayer IA, Diamond JR, et 37. Flaherty KT, Infante JR, Daud A, et al.
negative (triple-negative) breast cancer. al. Efficacy and safety of anti-Trop-2 anti- Combined BRAF and MEK inhibition in
Breast Cancer Res Treat 2010;121:261-71. body drug conjugate sacituzumab govite- melanoma with BRAF V600 mutations.
15. Park IH, Im SA, Jung KH, et al. Ran- can (IMMU-132) in heavily pretreated pa- N Engl J Med 2012;367:1694-703.
domized open label phase III trial of irino- tients with metastatic triple-negative breast 38. Chuk MK, Chang JT, Theoret MR, et
tecan plus capecitabine versus capecitabine cancer. J Clin Oncol 2017;35:2141-8. al. FDA approval summary: accelerated
monotherapy in patients with metastatic 26. Cortes J, O’Shaughnessy J, Loesch D, approval of pembrolizumab for second-
breast cancer previously treated with an- et al. Eribulin monotherapy versus treat- line treatment of metastatic melanoma.
thracycline and taxane: PROCEED trial ment of physician’s choice in patients with Clin Cancer Res 2017;23:5666-70.
(KCSG BR 11-01). Cancer Res Treat 2019; metastatic breast cancer (EMBRACE): 39. Beaver JA, Howie LJ, Pelosof L, et al.
51:43-52. a phase 3 open-label randomised study. A 25-year experience of US Food and Drug
16. Twelves C, Awada A, Cortes J, et al. Lancet 2011;377:914-23. Administration accelerated approval of ma-
Subgroup analyses from a phase 3, open- 27. Kawato Y, Aonuma M, Hirota Y, Kuga lignant hematology and oncology drugs
label, randomized study of eribulin me- H, Sato K. Intracellular roles of SN-38, a and biologics: a review. JAMA Oncol 2018;
sylate versus capecitabine in pretreated pa- metabolite of the camptothecin derivative 4:849-56.
tients with advanced or metastatic breast CPT-11, in the antitumor effect of CPT-11. Copyright © 2019 Massachusetts Medical Society.
cancer. Breast Cancer (Auckl) 2016;10:77- Cancer Res 1991;51:4187-91.
84. 28. Ocean AJ, Starodub AN, Bardia A, et


Bardia 2019

Uploaded by

Copyright:

Available Formats

Bardia 2019

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Bardia 2019

Uploaded by

Copyright:

Available Formats

The n e w e ng l a n d j o u r na l of m e dic i n e

Sacituzumab Govitecan-hziy in Refractory

on days 1 and 8 of each 21-day cycle until disease progression or unacceptable

n engl j med 380;8 nejm.org February 21, 2019 741

742 n engl j med 380;8 nejm.org February 21, 2019

The New England Journal of Medicine

n engl j med 380;8 nejm.org February 21, 2019 743

Table 2. Summary of Adverse Events in Patients Receiving Sacituzumab Govitecan-hziy.*

Adverse Event Patients (N=108)

Any Grade Grade 3 Grade 4

number of patients with event (percent)

744 n engl j med 380;8 nejm.org February 21, 2019

The New England Journal of Medicine

Adverse Event Patients (N=108)

Any Grade Grade 3 Grade 4

number of patients with event (percent)

n engl j med 380;8 nejm.org February 21, 2019 745

746 n engl j med 380;8 nejm.org February 21, 2019

The New England Journal of Medicine

Change from Baseline (%)

was 5.5 months (95% CI, 4.1 to 6.3); the estimated

umab govitecan-hziy with that of previous anti- 0.8

static triple-negative breast cancer for whom data 0.6

Discussion No. at Risk 108 73 43 22 12 7 5 3 3 1 1 1 0

Among patients with metastatic triple-negative

n engl j med 380;8 nejm.org February 21, 2019 747

Table 3. Summary of Treatment Efficacy, According to Local Assessment.

748 n engl j med 380;8 nejm.org February 21, 2019

The New England Journal of Medicine

Receipt of last previous therapy Receipt of sacituzumab govitecan-hziy Continuation of treatment

n engl j med 380;8 nejm.org February 21, 2019 749

750 n engl j med 380;8 nejm.org February 21, 2019

The New England Journal of Medicine

n engl j med 380;8 nejm.org February 21, 2019 751

You might also like