Safety and Efficacy of Bortezomib in AIDS-Associated Kaposi Sarcoma

Published OnlineFirst October 17, 2019; DOI: 10.1158/1078-0432.CCR-19-1044

CLINICAL CANCER RESEARCH | CLINICAL TRIALS: TARGETED THERAPY

Pilot Trial AMC-063: Safety and Efﬁcacy of Bortezomib in AIDS-associated Kaposi Sarcoma Erin G. Reid1, Adrienne Suazo1, Shelly Y. Lensing2, Dirk P. Dittmer3, Richard F. Ambinder4, Frank Maldarelli5, Robert J. Gorelick6, David Aboulaﬁa7, Ronald Mitsuyasu8, Mark A. Dickson9, and William Wachsman1,10; on behalf of the AIDS Malignancy Consortium (AMC)

ABSTRACT ◥ Purpose: AIDS-related Kaposi sarcoma is often incompletely 28-day cycles in patients with relapsed/refractory (r/r) AIDS- controlled, requiring serial therapies. Kaposi sarcoma herpesvirus Kaposi sarcoma taking antiretroviral therapy. (KSHV) induces transformation of endothelial cells, where it resides Results: Seventeen patients enrolled. No dose-limiting toxicities in a predominately latent state. We hypothesized proteasome occurred and the MTD was not reached. The most common adverse inhibition would have direct antitumor activity, induce lytic acti- events included diarrhea, fatigue and nausea. Among 15 evaluable vation of KSHV, and inhibit HIV infectivity, improving control of patients, partial response (PR) occurred in nine (60%), with a PR rate both Kaposi sarcoma and HIV. The primary objective was deter- of 83% in the 1.6 mg/m2 cohort; the remainder had stable disease (SD). mining the MTD of bortezomib in AIDS-Kaposi sarcoma. Second- Median time to response was 2.1 months. Median change in KSHV ary objectives included estimating the impact of bortezomib on PDCN was signiﬁcantly different between those with PR versus SD. Kaposi sarcoma response, KSHV plasma DNA copy number During cycle 1, seven of 11 evaluable patients had decreases in HIV VL. (PDCN), and HIV viral loads (VL). Conclusions: Bortezomib is well-tolerated and active in Patients and Methods: A3þ3 dose escalation design was AIDS-Kaposi sarcoma. The 60% PR rate is notable given the dose- employed evaluating four dose levels of bortezomib (0.75, 1, 1.2, or ﬁnding nature of the study in a r/r population. Changes in KSHV 1.6 mg/m2) administered intravenously on days 1, 8, and 15 of PDCN and HIV VL trended as hypothesized.

Introduction cutaneous disease. More extensive cutaneous or visceral disease requires systemic treatment, such as liposomal anthracyclines or Persons living with HIV (PLWH) are several thousand times more paclitaxel, which have overall response rates ranging from 40% to likely to be diagnosed with Kaposi sarcoma than their HIV seroneg- 90% (3, 5, 6). Small studies of targeted therapies and immunomod- ative counterparts, and Kaposi sarcoma is one of the most commonly ulatory agents demonstrating activity in AIDS-Kaposi sarcoma diagnosed malignancies among PLWH (1, 2). While the introduction include imatinib [n ¼ 30; overall response rate (ORR), 33%; ref. 7], of highly active antiretroviral therapy (HAART) has been associated thalidomide (n ¼ 17; ORR, 47%; ref. 8), and pomalidomide (n ¼ 15; with decreased incidence of and increased survival rates in Kaposi ORR, 60%; ref. 9). sarcoma, complete remission is rare and there is no curative therapy. Kaposi sarcoma–associated herpesvirus (KSHV, also known as Immune reconstitution through the use of HAART is an important human herpesvirus-8, HHV-8) is the primary etiologic factor associ- component of therapy, but is often insufﬁcient, especially with ated with Kaposi sarcoma. The virus infects endothelial cells, resulting advanced disease (3, 4). Local therapies such as radiation, cryotherapy, in transformation into spindle cells characteristic of Kaposi sarco- intralesional, or topical therapies may be used to control very limited ma (10). Residing in a predominately latent state within Kaposi sarcoma, KSHV promotes tumor cell survival. The exact mechanisms 1University of California, San Diego Moores Cancer Center, La Jolla, California. of KSHV-driven oncogenesis are unknown and may involve any 2University of Arkansas for Medical Sciences, Little Rock, Arkansas. 3Lineberger number of viral genes, such as the latency-associated nuclear antigen Comprehensive Cancer Center at the University of North Carolina at Chapel Hill, (LANA), the viral cyclin (vCYC) and a protein called vFLIP, which North Carolina. 4Johns Hopkins School of Medicine, Baltimore, Maryland. 5HIV inhibit apoptosis and autophagy, and activate NF-kB (11) as well as the Dynamics and Replication Program, National Cancer Institute, Frederick, viral miRNAs. Acting together, KSHV reprograms both infected and 6 Maryland. AIDS and Cancer Virus Program, Frederick National Laboratory for uninfected cells within the Kaposi sarcoma lesion (12). Cancer Research, Frederick, Maryland. 7Floyd and Delores Jones Cancer Institute at Virginia Mason Medical Center, Seattle, Washington. 8Center for AIDS Research Bortezomib (Velcade, Millennium Pharmaceuticals, Inc.) is a pro- and Education, University of California, Los Angeles, Los Angeles, California. teasome inhibitor that was initially approved for treatment of multiple 9Memorial Sloan Kettering Cancer Center and Weill Cornell Medical College, New myeloma and has activity in other malignancies attributed, at least in York, New York. 10VA San Diego Healthcare System, San Diego, California. part, to inhibition of NF-kB (13, 14–16). We hypothesized bortezomib Note: Supplementary data for this article are available at Clinical Cancer would be active in Kaposi sarcoma either as a result of inhibition of Research Online (http://clincancerres.aacrjournals.org/). proteasome function (required for NF-kB) or as a result of KSHV lytic ClinicalTrials.gov ID NCT01016730 activation resulting in direct tumor killing by viral-mediated lysis. Antitumor activity of bortezomib against Kaposi sarcoma–endothelial Corresponding Author: Erin G. Reid, Moores Cancer Center, University of California, San Diego, 3855 Health Sciences Drive, La Jolla, CA 92093-0987. cell models and PEL has been noted in preclinical studies (17, 18). Phone: 858-822-6271; Fax: 858-822-6288; E-mail: [email protected] Similarly, bortezomib-mediated viral lytic gene expression has been seen in cell line xenograft studies.(19) Bortezomib has been identiﬁed Clin Cancer Res 2020;26:558–65 as a potent lytic activator of Epstein Barr virus (EBV), a gammaher- doi: 10.1158/1078-0432.CCR-19-1044 pesvirus similar to KSHV (20, 21), as well as KSHV in the context of the 2019 American Association for Cancer Research. KSHV-associated primary effusion lymphoma (PEL; refs. 22–24).

AACRJournals.org | 558

Safety and Efﬁcacy of Bortezomib in AIDS-associated KS

This study was conducted in accordance with the Declaration of Translational Relevance Helsinki. Institutional review boards at each of the study sites approved This trial is the first to demonstrate clinical activity of protea- the study protocol in accordance with an assurance filed with and some inhibition in the management of Kaposi sarcoma. Admin- approved by the U.S. Department of Health and Human Services. All istration of bortezomib was safe with only mild-to-moderate patients provided written informed consent prior to study procedures adverse effects and showed excellent activity in AIDS-related and treatment. Kaposi sarcoma with 100% disease control rate and partial Required supportive care included chemoprophylaxis against her- þ responses occurring in five of six patients treated at the highest pes zoster, pneumocystis jiroveci pneumonia for patients with CD4 dose level. Compared with other currently utilized treatments for counts <200 cells/mm3, and Mycobacterium-avium complex (MAC) þ relapsed/refractory AIDS-related Kaposi sarcoma, bortezomib for patients with CD4 counts <50 cells/mm3. G-CSF was not required demonstrated an equivalent or greater response rate, with few for neutropenia prophylaxis, but was permitted for treatment when required dose reductions or delays and lack of alopecia or skin clinically indicated. The dose-limiting toxicity (DLT) period was toxicities. Despite concerns about potential risks of lytic Kaposi defined as two cycles (8 weeks). Patients who tolerated bortezomib sarcoma herpesvirus activation, bortezomib was well-tolerated and without experiencing any DLTs and had at least stable disease or better neither tumor progression nor other indications of adverse events after two cycles were allowed to continue on protocol therapy for a attributable to viral lytic activation were seen. Given that the maximum of eight cycles. Intrapatient dose escalation to the next dose highest burden of Kaposi sarcoma falls upon resource-limited level was permitted in patients without DLTs after cycle 2. countries, evaluation of strategies employing orally administered Given the theoretical risk of inflammatory complications of lytic proteasome inhibitors or combination therapies including protea- viral replication, patients were monitored clinically for development of some inhibitors in Kaposi sarcoma would be attractive. life-threatening inflammatory syndromes (including multicentric Castleman disease, severe immune reconstitution syndrome, or KSHV inflammatory cytokine syndrome); the occurrence of any such event would require discontinuation of protocol therapy. Monitoring for Lytic activation of the latent KSHV infecting Kaposi sarcoma tumors such inflammatory syndromes included regular clinical examination may result not only in direct tumor cell lysis but also in increased utilizing history and vital signs to evaluate for new onset fatigue, antigen expression and immune recognition of Kaposi sarcoma, as dyspnea, myalgias, arthralgias or fevers; physical exam to evaluate for lytic viral proteins are often more immunogenic than their latent progression of cutaneous Kaposi sarcoma, lymphadenopathy or devel- counterparts. Alternatively, lytic activation may conceivably result in opment of effusions; and laboratory investigations to assess for viral inflammatory syndromes such as multicentric Castleman disease cytopenias (complete blood count with differential), hyponatremia or KSHV-associated inflammatory cytokine syndrome (KICS), or even or organ failure (complete metabolic panel), or inflammation drive tumorigenesis. (C-reactive protein). Chest imaging was required at baseline and In addition to the inhibitory effect of bortezomib on tumors latently at onset of pulmonary symptoms to evaluate for progressive Kaposi infected by gammaherpesviruses, preclinical models suggest that sarcoma, development of effusions, or infectious complications. bortezomib may suppress HIV via the cytidine deaminase, APO- BEC3G, an innate cellular human defense against retroviruses. Human Key eligibility criteria APOBEC3G is incorporated into nascent retroviral virions as they bud Inclusion criteria included confirmation of HIV seropositivity with from the host cell and hypermutates viral cDNA within newly infected an FDA-approved and licensed test. Kaposi sarcoma diagnosis had to cells, thereby impairing faithful viral replication (Fig. 1A). HIV evades be confirmed by biopsy and relapsed after or refractory to at least one this defense by employing virion infectivity factor (Vif), which induces other prior first-line systemic therapy, either liposomal doxorubicin or the ubiquitination of APOBEC3G, and targets the protective protein paclitaxel. Individuals must have had at least five measurable cuta- for proteasomal degradation (Fig. 1B). In vitro studies inhibiting neous Kaposi sarcoma lesions suitable for assessment in addition to human 26S proteasome have demonstrated impairment of HIV lesions for baseline and on study biopsies. Patients were required to be viral budding and infectivity in association with restoration of host on a stable (12 weeks) protease inhibitor–based or non-nucleotide APOBEC3G levels (Fig. 1C; refs. 25, 26). We recently reported an reverse transcriptase inhibitor–based ART regimen consisting of at increase in APOBEC3G levels in patients with HIV lymphoma treated least three drugs, with no intention of changing the regimen once on with bortezomib (27), suggesting bortezomib has the potential to study. Other key inclusion criteria included: age 18 years, absolute contribute to suppression of HIV through restoration of this protective neutrophil count 1,000/mm3, hemoglobin 8.0 g/dL, platelet count factor. Given the important role immune reconstitution plays in 100,000/mm3, total bilirubin 1.5 mg/dL, AST/ALT 2 ULN, and control of AIDS-Kaposi sarcoma, this additional antiretroviral activity creatinine clearance 50 mL/minute/1.73 m2 as calculated by the may be beneficial. Cockcroft–Gault equation. Patients were excluded if they had any improvement of Kaposi sarcoma during the 4 weeks prior to enrollment, or if they had symptomatic visceral or pulmonary Kaposi Patients and Methods sarcoma. Patients with oral and/or lymph node involvement, and/or Protocol design and therapy asymptomatic pulmonary Kaposi sarcoma not interfering with activ- A3þ3 dose escalation design of single-agent bortezomib was used ities of daily living were eligible. Other key exclusion criteria included to determine the MTD of bortezomib and to evaluate the clinical an Eastern Cooperative Oncology Group performance score >2, grade response of Kaposi sarcoma tumors to bortezomib. Bortezomib was >2 baseline peripheral neuropathy, concurrent active opportunistic administered intravenously on days 1, 8, and 15 of each 28-day cycle at infection, concurrent additional malignancy, serious medical illness one of four dose levels: 0.75 mg/m2, 1 mg/m2, 1.2 mg/m2, and 1.6 mg/ requiring acute therapy within 14 days prior to study entry, hyper- m2, respectively. sensitivity to Boron, or expected survival < 3 months.

AACRJournals.org Clin Cancer Res; 26(3) February 1, 2020 559

Reid et al.

B A Human defense against retrovirus Vif protects HIV infections: APOBEC3G against APOBEC3G

No viral replication

APOBEC3G incorporated into APOBEC3G hypermutates viral Degradation of ubiquinated virion as it buds from cDNA during reverse transcription, HIV replicates in APOBEC3G by proteasome infected cell blocking viral replication in newly and buds from infected cell newly infected cell

C Proposed anti-HIV activity of bortezomib

Key APOBEC3G No HIV Retroviral RNA replication Retroviral cDNA Mutated retroviral cDNA Vif Bortezomib inhibits proteasome HIV replication blocked Ubiquitin-tagging degradation of ubiquitinated in newly infected cell APOBEC3G allowing its Proteasome accumulation and incorporation into Bortezomib budding virion

Figure 1. Mechanism of hypothesized anti-retroviral effects of bortezomib. Humans have an innate defense mechanism against retroviral infections employing a cytidine deaminase, APOBEC3G. A, APOBEC3G is incorporated into nascent retroviral virions as they bud from an infected cell. As the nascent virion infects a cell and begins transcribing and copying its genes into the cellular DNA, APOBEC3G hypermutates the viral DNA to a point where replication is impossible, breaking the chain of infection. B, HIV circumvents this antiviral pathway using one of its accessory gene products, virion infectivity factor (VIF). Virion infectivity factor induces ubiquitinization of APOBEC3G, thereby targeting this factor for proteasome degradation. In this way, VIF effectively shortens the half-life of APOBEC3G within the cell and prevents its incorporation into new virions. C, Proteasome inhibition would be expected to reverse the effects of VIF by slowing the degradation of APOBEC3G allowing it to be incorporated into budding HIV virions, impairing faithful viral replication.

Measurements and assays Methods). The Functional Assessment of Cancer Therapy/Gynecologic AIDS Clinical Trials Group (ACTG) Kaposi sarcoma Staging Oncology Group–Neurotoxicity (FACT-GOG/NTX) question- Classification was implemented to determine staging of disease at naire (31) was administered at baseline, day 15 of cycle 1 and the study entry. Two 3-mm punch biopsies were collected from a Kaposi first cycle of intrapatient dose escalation, and on the fourth week of sarcoma lesion at baseline, cycle 1 day 2, and at treatment discontin- every subsequent treatment cycle. uation. One of the two biopsies was submerged in RNAlater for RNA isolation, while the other was fixed in formalin for further assessment Statistical analysis by IHC. Descriptive statistics were reported for demographic data, baseline IHC from biopsy specimens was performed to evaluate tumor disease characteristics, adverse events, and Kaposi sarcoma response content and proliferation (by Ki-67), and presence of KSHV (LANA). rates. Binomial proportions with 95% confidence intervals were Messenger RNA was isolated from RNA-later preserved biopsies using calculated to estimate the overall Kaposi sarcoma response rates for real-time reverse-transcription quantitative polymerase chain reaction all patients considered evaluable for response. The Wilcoxon signed (qRT-PCR) to measure the levels of all KSHV genes in the tumor, as rank test was used to investigate the changes from pre- to postborte- described previously (28). zomib administration on HIV viral loads from serum, on KSHV copy Blood samples were collected at baseline, cycle 1 day 1, 2, 4, 8, 15 and number from Kaposi sarcoma, plasma, and FACT-GOG/NTX scores. cycle 2 day 1. KSHV PCR on DNA extracted from plasma was carried Unsupervised clustering and principal component analysis were out as described previously (29). Plasma HIV viral load was deter- applied to explore KSHV transcription in Kaposi sarcoma lesions. mined using qRT-PCR, as described previously (30). HIV plasma samples were also collected on day 1 of subsequent cycles prior to bortezomib administration. Results Tumor response assessments were performed at baseline, weeks 4, 8, Seventeen patients were enrolled across six different sites within the 12, 16, and 20, and then every 3 months for up to one year using AIDS Malignancy Consortium (AMC). As shown in Table 1, all of the response definitions outlined in the protocol (see Supplementary patients were males with a median age (range) of 45 (29–59) years. At

560 Clin Cancer Res; 26(3) February 1, 2020 CLINICAL CANCER RESEARCH

Safety and Efﬁcacy of Bortezomib in AIDS-associated KS

Table 1. Patient characteristics (n ¼ 17). in the context of endocarditis without associated neutropenia or lymphopenia; his endocarditis was attributed to a PICC line infection Demographics occurring in the context of performing gardening and heavy land- Age years, median (range) 45 (29–59) scaping work without adequate protection of the PICC line from Gender, n (%) exposure. Male 17 (100) Grade 1 peripheral neuropathy was reported in four patients (24%), Race/ethnicity, n (%) and an additional patient developed grade 2 neuralgia. Neurotoxicity as Non-Hispanic, white 8 (47) measured by the FACT/GOG-Neurotoxicity Questionnaire did not Non-Hispanic, black/AA 3 (18) significantly increase at any assessment point in this small study. With Non-Hispanic, multiracial 2 (12) apossiblescorerangeof0–44, the median baseline score was 6. The Hispanic, white 3 (18) median change from baseline was 0.5 [interquartile range (IQR), 1.5 Unknown/not reported 1 (6) to 6.0] by treatment discontinuation (n ¼ 12, P ¼ 0.402). The maximum Staging at registration Tumor stage, n (%) score increase was 10 points, occurring in two patients, corresponding to 1. Tumor is confined to skin and/or to lymph nodes and/or 10 (59) development of grade 1 peripheral sensory neuropathy. patient has minimal oral disease Of303 planned doses,89% were administered on time at the protocol- 2. Tumor with edema, ulceration, or extensive oral KS or 7(41) specified dose. There were 21 dosing delays of which 11 (52%) were due gastro KS or KS in other non-nodal viscera to an adverse event and others were due to patient preference or Immune System Staging, n (%) scheduling. The median delay was 7 days. There were four dose CD4 cells 200/mL 15 (88) reductions, two of which were due to adverse events, one due to < m CD4 cells 200/ L 2 (12) physician error, and the other due to patient request. Two patients History of thrush/opportunistic infection, n (%) discontinued due toadverse events atleast possibly related to bortezomib Yes 7 (41) fi No 10 (59) (grade 3 neutropenia considered de nitely related but would not have Karnofsky performance status, n (%) required protocol discontinuation, and grade 2 lymphedema considered <70 3 (18) possibly related). 70 14 (82) Among the 15 patients that were evaluable for response, 60% (95% HIV viral load, n (%) CI, 32%–84%) experienced a partial response (PR) as their best Detectable 4 (24) response to therapy, whereas the remaining patients had stable disease Undetectable 13 (76) (SD; Table 3). Five of 6 patients treated at the highest dose level experienced a PR (83%). The median time to response was 2.1 (IQR, 1.9–2.8) months. Examples of lesion improvement as early as post baseline, 59% (n ¼ 10) of patients had disease involvement confined to cycle 2 are included in Supplementary Materials. By the end of the the skin (without associated edema or ulcerations) and/or to lymph study, none of the patients who experienced a PR progressed after a nodes and/or had minimal oral disease. All but two of the 17 patients median of 2.8 (IQR, 1.9–18.5) months from initial response. There was þ had a CD4 count >200 cell/mL (88%, n ¼ 15). No patient had active no disease progression while on protocol therapy. One patient with SD organ or visceral involvement with Kaposi sarcoma. Fourteen patients progressed 446 days after enrollment. (82%) had a Karnofsky Performance Status (KPS) of >70. Thirteen Median change in plasma KSHV DNA copy number during cycle 1 patients (76%) had an undetectable HIV viral load at baseline. compared with baseline was negligible in patients obtaining partial re- Three patients were enrolled into each of the four dose escalation mission. In contrast, the median plasma KSHV DNA copy number cohorts of the study without occurrence of a DLT. The MTD was not decreased compared with baseline in patients with stable disease. The reached at the highest dose, 1.6 mg/m2, which was selected for a dose difference between responders and nonresponders was statistically signi- expansion cohort to which an additional three patients were enrolled. ficant after the first treatment on both cycle 1 day 2 and day 8 (Fig. 2). In Of the three patients, two withdrew early from the study prior to the cohort treated at the highest dose level of bortezomib, seven of eight completing two cycles of therapy so were not evaluable for response patients were evaluable for changes in KSHV plasma DNA copy number and were replaced. The first patient declined continued protocol during cycle 1. Of these seven, six patients had increases (range of maxi- therapy and follow-up, and the second withdrew from the trial due mum increase: 71–40,517 copies/mL) while the other patient's KSHV was to scheduling difficulties. A total of six patients treated at the highest persistently undetectable. The timing of the maximum increase in KSHV dose level were therefore evaluable for response. One patient in cohort copy number during the first cycle was variable among patients. 1 and two patients in cohort 2 dose escalated to the next higher dose The numbers of tumor cells in KS lesions as assessed by IHC and level for their final two cycles of protocol therapy. mRNA profiling varied widely among patients. Figure 3 shows a Of the 17 patients enrolled and treated on study therapy, all of them heatmap depiction of relative mRNA levels for KSHV and indicator experienced at least one adverse event (AE). There were a total of 177 endothelial cell (EC) mRNAs. Darker red colors indicate more and AEs experienced, the most common of which were diarrhea (53%), lighter yellow colors less transcription. The data were adjusted for fatigue (53%), and nausea (41%). Nearly two-thirds of all AEs (n ¼ equal mRNA levels using a combination of multiple “housekeeping” 112/177) were attributed as at least possibly related to bortezomib, and genes. The mRNAs clustered into three broad categories: “EC,” which were experienced by 15 (88%) of participants (Table 2). is comprised of human EC markers, “latent,” which comprises KSHV Three patients experienced serious adverse events (SAEs). One latent mRNAs, that should be present in every infected cell, and “lytic,” patient had severe vomiting. A second patient had two separate which is comprised of KSHV mRNAs that are associated with lytic infectious SAEs, both occurring without associated neutropenia or reactivation in PEL cell lines. The biopsies clustered into three clusters lymphopenia: bacteremia assessed as grade 4 sepsis in the context of “a,”“b,” and “c.” Cluster “c” represents lesions with high-level KSHV sinusitis and pneumonic process, and a grade 3 cellulitis of the thigh at activity, cluster “b” represents lesions with a large proportion of ECs, site of Kaposi sarcoma. A third patient experienced fatal cardiac arrest but KSHV was present in only a minority of EC and predominantly in a

AACRJournals.org Clin Cancer Res; 26(3) February 1, 2020 561

Reid et al.

Table 2. Adverse events.

Severity grade total number of occurrences (number attributed as at least possibly related to bortezomib) Life- Organ system/adverse event Mild Moderate Severe threatening Death Total

Blood and lymphatic system disorders Anemia 3 (3) 1 (1) 4 (4) Cardiac disorders Cardiac arrest 11 Sinus tachycardia 1 (1) 1 (1) Ear and labyrinth disorders Hearing impaired 1 1 Eye disorders Blurred vision 1 2 3 Conjunctivitis 1 1 Gastrointestinal disorders Abdominal pain 1 (1) 1 (1) 2 (2) Constipation 2 (2) 2 (2) Diarrhea 7 (5) 6 (5) 13 (10) Dry mouth 1 (1) 1 (1) Gastroesophageal reﬂux disease 2 (2) 2 (2) Mucositis oral 1 (1) 1 (1) Nausea 5 (5) 6 (5) 1 (1) 12 (11) Vomiting 4 (3) 2 (2) 1 (1) 7 (6) General disorders and administration Chills 4 (2) 4 (2) site conditions Edema limbs 1 (1) 1 2 (1) Fatigue 8 (7) 2 (2) 10 (9) Fever 5 (4) 1 (1) 6 (5) Malaise 1 (1) 1 (1) Noncardiac chest pain 1 1 2 Pain 4 4 Infections and infestations Endocarditis infective 1 1 Infections and infestations—other, 11 specify Sepsis 1 1 Sinusitis 3 3 Skin infection 2 (1) 2 1 5 (1) Upper respiratory infection 2 (1) 2 (1) Investigations Alanine aminotransferase increased 1 (1) 1 (1) Blood bilirubin increased 1 1 Cholesterol high 1 (1) 1 (1) Lymphocyte count decreased 1 (1) 1 (1) Neutrophil count decreased 2 (2) 3 (3) 5 (5) Platelet count decreased 4 (3) 4 (3) White blood cell decreased 5 (5) 1 (1) 6 (6) Metabolism and nutrition disorders Hyperglycemia 3 3 Hyperkalemia 1 1 Hypocalcemia 5 (5) 5 (5) Hyponatremia 7 (6) 1 8 (6) Musculoskeletal and connective tissue Arthralgia 1 1 disorders Back pain 1 (1) 2 1 4 (1) Muscle weakness lower limb 1 (1) 1 (1) Myalgia 1 (1) 1 (1) Pain in extremity 3 1 4 Nervous system disorders Dizziness 2 (2) 2 (2) Headache 5 (4) 2 (1) 7 (5) Memory impairment 2 2 Nervous system disorders—other 2 (1) 2 (1) Peripheral motor neuropathy 2 (1) 2 (1) Peripheral sensory neuropathy 2 (1) 1 (1) 3 (2) Tremor 1 1 Psychiatric disorders Anxiety 1 (1) 1 (1) Depression 3 3 Renal and urinary disorders Acute kidney injury 1 (1) 1 (1) Respiratory, thoracic and mediastinal Cough 1 (1) 1 (1) 2 (2) disorders Dyspnea 1 1 2 Pneumonitis 1 (1) 1 (1) Sinus disorder 1 (1) 1 (1) (Continued on the following page)

562 Clin Cancer Res; 26(3) February 1, 2020 CLINICAL CANCER RESEARCH

Safety and Efﬁcacy of Bortezomib in AIDS-associated KS

Table 2. Adverse events. (Cont'd ) Severity grade total number of occurrences (number attributed as at least possibly related to bortezomib) Life- Organ system/adverse event Mild Moderate Severe threatening Death Total

Skin and subcutaneous tissue disorders Hyperhidrosis 1 (1) 1 (1) Skin and subcutaneous tissue disorders- 22 other Vascular disorders Hypertension 1 1 Lymphedema 2 (2) 2 (2) Superﬁcial thrombophlebitis 1 1 Total (Occurrences) 95 (72) 65 (34) 14 (6) 2 1 177 (112)

Note: Total adverse events occurrences are presented by category and grade followed in parentheses by the number of occurrences attributed as at least possibly related to bortezomib. latent state. Finally, cluster “b” represents lesions with virtually no mib was standard of care at the time of protocol design, subcutaneous endothelial cells and no KSHV, that is, fibrotic tissue rather than tumor administration was not employed, but would be expected to further tissue. The IHC analyses confirm the profiling data. Every biopsy reduce risk of neuropathy (32). stained positive for Ki-67, had clusters of KSHV LANA-positive Changes in HIV viral loads were not statistically significant; how- spindle cells, but often these clusters consisted of very few cells. ever, there was a trend toward decreased HIV viral loads during cycle 1 Of 11 patients evaluable for change in HIV VL from baseline during as predicted. The small size of the study and undetectable baseline HIV cycle 1, seven had a decrease (range of maximum absolute decrease: viral loads in 76% of patients limited sensitivity and interpretation of 2.5–1,330 copies/mL), and four had an increase in HIV VL (range of the data despite single copy viral load testing. maximum absolute increase: 0.8 to 54 copies/mL). Of the five patients KSHV copy number in plasma varied considerably from patient to with >1 log decrease in HIV VL during cycle 1, four experienced PR patient. In general, viral DNA persisted in plasma throughout the and the other had SD. course of the study. There was a trend toward increased KSHV plasma DNA in patients receiving the highest dose of bortezomib (1.6 mg/m2), an expected finding with hypothesized lytic KSHV activation. How- Discussion ever, it is not clear whether this reflected infectious virion production Administration of bortezomib was safe and well-tolerated, with only as might be expected from a lytic activator or merely the release of viral mild-to-moderate adverse effects attributed to bortezomib, and DNA from dying Kaposi sarcoma cells. We note that no inflammatory showed excellent activity in AIDS-Kaposi sarcoma with 100% disease syndromes were observed in the trial, a potential risk of viral lytic control rate. No DLTs occurred at the highest tested dose level at which activation for which the trial specified monitoring. No patient with all but one patient experienced a partial response. Recognizing this is a stable disease had increase in KSHV copy number during cycle 1 and phase I study with small numbers, comparisons with other therapies the median plasma KSHV DNA copy number decreased compared are inherently limited. However, bortezomib demonstrated a high with baseline in patients with stable disease. The explanation for this proportion of responses in a population refractory to standard first- trend in nonresponding patients is under investigation. line therapy of Kaposi sarcoma, with few cytopenias, a low proportion Mechanistic and molecular studies on tumor tissue were difficult to of dose reductions or delayed treatment, and lack of alopecia. interpret due to the small number of Kaposi sarcoma spindle cells in the Increased risk of infections and neuropathy are known complica- biopsies at baseline. The presence of two types of Kaposi sarcoma tions of both bortezomib and HIV infection. Severe infections were differing by the degree of Kaposi sarcoma transcription recapitulates observed in the trial and assessed as unlikely related to study therapy prior studies of Kaposi sarcoma from ART-na€ve patients in Sub-Saharan due to alternate attributions. However, given the well-documented Africa and AIDS-Kaposi sarcoma patients in the United States (28, 33). increased risk of infections due to proteasome inhibitor treatment in The fact that over half of the biopsies exhibited minimal EC and Kaposi other disease contexts, vigilant monitoring remains important when sarcoma transcription may reflect the patient composition of AMC-063, using proteasome inhibition in this population. Neurotoxicity was not which enrolled only patients on successful ART in the United States. found to be a limiting adverse event of this bortezomib dosing scheme Multiple factors limited sensitivity to changes in KSHV viral gene in AIDS-Kaposi sarcoma. As intravenous administration of bortezo- expression: on-treatment biopsies were limited to a single assessment

Table 3. Best responsea.

Bortezomib Bortezomib Bortezomib Bortezomib All 0.75 mg/m2 1.0 mg/m2 1.3 mg/m2 1.6 mg/m2 doses N ¼ 3 N ¼ 3 N ¼ 3 N ¼ 6 N ¼ 15

Partial response, n (%) 1 (33) 3 (100) 0 (0) 5 (83) 9 (60) Stable disease, n (%) 2 (67) 0 (0) 3 (100) 1 (17) 6 (40) aTwo patients (not shown above) were not evaluable for response as they did not complete two cycles of therapy. Intra-patient dose escalation to the next dose level was permitted in patients without DLTs after cycle 2. One patient in cohort 1 and two patients in cohort 2 dose-escalated to the next higher dose level for their ﬁnal two cycles of protocol therapy.

AACRJournals.org Clin Cancer Res; 26(3) February 1, 2020 563

Reid et al.

direct actions of bortezomib on cellular pathways independent of KSHV, perhaps through induction of an unfolded protein response, as well as actions of bortezomib on KSHV reactivation from latency, which lead to cell death. Reactivating KSHV from latency comes with a theoretical risk as KSHV lytic proteins can have protumorigenic, proinflammatory and immune-suppressive phenotypes (35), and as lytic replication may foster further dissemination of the virus. Inducers of KSHV reactivation without direct cellular toxicity have demonstrated limited efficacy in Kaposi sarcoma (36). The combination of direct cellular cytotoxicity and potential to reactivate KSHV found in bortezomib may have contributed to the clinical efficacy of bortezomib and did not result in KSHV-associated inflammatory syndromes. The study is somewhat limited given the small sample size and relatively short median follow- up. Although prophylaxis with valganciclovir to protect against KSHV activation has been suggested, its use is controversial, and in the patients Figure 2. treated in this trial without such prophylaxis, no adverse effects Change in KSHV DNA copy number from baseline for cycle 1 according to best attributable to KSHV viremia were identified. Furthermore, there were response. Change in plasma KSHV DNA copy number during cycle 1 was no cases of tumor progression during bortezomib therapy (37, 38). compared with baseline. The median change (represented by horizontal lines) While the precise mechanism of action of bortezomib as an was negligible in patients obtaining partial remission (individual patient values antitumor agent in Kaposi sarcoma remains unclear, the data pre- represented as dots). In contrast, the median plasma KSHV DNA copy number decreased compared with baseline in patients with stable disease (individual sented demonstrate that bortezomib is active in patients who have patient values represented as squares). The difference between responders and failed standard therapies. Despite concerns about potential risks of nonresponders was statistically significant after the first treatment on both cycle lytic KSHV activation, bortezomib was well tolerated and neither 1 day 2 and day 8. tumor progression nor other indications of adverse events attributable to viral lytic activation were seen. Therefore, further evaluation of performed only 24 hours after initial bortezomib exposure, not all proteasome inhibition is warranted. Given that the highest burden of baseline biopsies had robust evidence of KSHV transcription, and we Kaposi sarcoma falls upon resource-limited countries, evaluation of were not able to biopsy the same lesion multiple times. strategies employing an orally administered proteasome inhibitor or Bortezomib has multiple, divergent effects in KSHV-associated combination therapies including proteasome inhibitors in Kaposi malignancies. The bulk of mechanistic studies have been conducted sarcoma would be attractive. On the basis of the encouraging results in the KSHV-associated primary effusion lymphoma, which is depen- of this study, the AIDS Malignancy Consortium is developing a trial dent on NF-kB signaling (34). Investigations in KSHV-driven endo- using an oral proteasome inhibitor in Kaposi sarcoma. thelial cell models are more limited (17). The molecular mechanisms fl and specific targets underlying the efficacy of bortezomib in Kaposi Disclosure of Potential Con icts of Interest fl sarcoma are still unknown. The most likely scenario includes both No potential con icts of interest were disclosed. Disclaimer The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply the endorsement by the U.S. government. Authors’ Contributions Conception and design: S.Y. Lensing, R.F. Ambinder, F. Maldarelli, D.M. Aboulafia, W. Wachsman, E.G. Reid Development of methodology: R.F. Ambinder, D.P. Dittmer, W. Wachsman, E.G. Reid Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): A. Suazo, D.P. Dittmer, R.F. Ambinder, F. Maldarelli, R.J. Gorelick, D.M. Aboulafia, R. Mitsuyasu, M.A. Dickson, E.G. Reid Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): A. Suazo, S.Y. Lensing, D.P. Dittmer, R.F. Ambinder, F. Maldarelli, W. Wachsman, E.G. Reid Writing, review, and/or revision of the manuscript: E.G. Reid, A. Suazo, S.Y. Lensing, D.P. Dittmer, R.F. Ambinder, R.J. Gorelick, D.M. Aboulafia, Figure 3. R. Mitsuyasu, M.A. Dickson, W. Wachsman Relative mRNA levels for KSHV and indicator endothelial cell mRNAs. Shown is a Administrative, technical, or material support (i.e., reporting or organizing data, heatmap depiction of relative mRNA levels for KSHV and indicator endothelial constructing databases): A. Suazo, S.Y. Lensing cell (EC) mRNAs. Darker red colors indicate more and lighter yellow colors less Study supervision: W. Wachsman, E.G. Reid transcription. Clusters of samples are indicated by letters “a,”“b,” and “c.” One of 11 (10%) of biopsies taken day 2 of cycle 1 (gray) had evidence of KSHV transcription (asterisk), nine (90%) biopsies taken day 2 of cycle 1 showed no Acknowledgments molecular evidence of KSHV or EC (cluster “b”). Nine of 18 (50%) of baseline We gratefully acknowledge the participating patients and their caregivers, as well biopsies and three of nine (33%) biopsies taken at treatment discontinuation as all AMC subinvestigators and research staff contributing to the conduct of this were classified as KSHV/EC transcription positive. protocol. This study was supported by NCI U01 AMC grant (UM1CA121947), NIH

564 Clin Cancer Res; 26(3) February 1, 2020 CLINICAL CANCER RESEARCH

Safety and Efﬁcacy of Bortezomib in AIDS-associated KS

Bench to Bedside Award (3U01CA121947-03S2; to E.G. Reid and F. Maldarelli), NIH The costs of publication of this article were defrayed in part by the payment of page Cancer Center Support Grant (NCI CCC 5P50 CA23100-25): UCSD Cancer Clinical charges. This article must therefore be hereby marked advertisement in accordance Investigator Team Leadership Award (to E.G. Reid), UCLA Center for AIDS Research with 18 U.S.C. Section 1734 solely to indicate this fact. (P30 AI028697; to R. Mitsuyasu), UCLA Clinical Translational Sciences Institute (UL1 TR0001881; R. Mitsuyasu), NIDCR RO1 DE018304 (to D.P. Dittmer), and PHS grant CA019014 (to D.P. Dittmer). This work was also supported in part with federal Received March 28, 2019; revised June 10, 2019; accepted October 14, 2019; funds from the NCI, NIH, under contract No. HHSN261200800001E. published ﬁrst October 17, 2019.

References 1. NastiG,TalaminiR,AntinoriA,Martellotta F, Jacchetti G, Chiodo F, et al. 19. Sarosiek KA, Cavallin LE, Bhatt S, Toomey NL, Natkunam Y, Blasini W, et al. AIDS-related Kaposi's Sarcoma: evaluation of potential new prognostic Efficacy of bortezomib in a direct xenograft model of primary effusion lym- factors and assessment of the AIDS Clinical Trial Group Staging System in phoma. Proc Natl Acad Sci U S A 2010;107:13069–74. the HAART era–the Italian Cooperative Group on AIDS and Tumors and the 20. Fu DX, Tanhehco YC, Chen J, Foss CA, Fox JJ, Lemas V, et al. Virus-associated tumor Italian Cohort of Patients Naive From Antiretrovirals. J Clin Oncol 2003;21: imaging by induction of viral gene expression. Clin Cancer Res 2007;13:1453–8. 2876–82. 21. Shirley CM, Chen J, Shamay M, Li H, Zahnow CA, Hayward SD, et al. 2. Grulich AE, van Leeuwen MT, Falster MO, Vajdic CM. Incidence of cancers in Bortezomib induction of C/EBPb mediates Epstein-Barr virus lytic activation people with HIV/AIDS compared with immunosuppressed transplant recipi- in Burkitt lymphoma. Blood 2011;117:6297–6303. ents: a meta-analysis. Lancet 2007;370:59–67. 22. Brown HJ, McBride WH, Zack JA, Sun R. Prostratin and bortezomib are novel 3. Martin-Carbonero L, Barrios A, Saballs P, Sirera G, Santos J, Palacios R, et al. inducers of latent Kaposi's sarcoma-associated herpesvirus. Antivir Ther 2005; Pegylated liposomal doxorubicin plus highly active antiretroviral therapy versus 10:745–51. highly active antiretroviral therapy alone in HIV patients with Kaposi's sarcoma. 23. Granato M, Romeo MA, Tiano MS, Santarelli R, Gonnella R, Gilardini Montani AIDS 2004;18:1737–40. MS, et al. Bortezomib promotes KHSV and EBV lytic cycle by activating JNK and 4. Hosseinipour MC, Kang M, Krown SE, Bukuru A, Umbleja T, Martin JN, et al. autophagy. Sci Rep 2017;7:13052. As-needed vs immediate etoposide chemotherapy in combination with antire- 24. Fu DX, Tanhehco Y, Chen J, Foss CA, Fox JJ, Chong JM, et al. Bortezomib- troviral therapy for mild-to-moderate AIDS-associated Kaposi sarcoma in induced enzyme-targeted radiation therapy in herpesvirus-associated tumors. resource-limited settings: A5264/AMC-067 randomized clinical trial. Clin Infect Nat Med 2008;14:1118–22. Dis 2018;67:251–60. 25. Liu B, Yu X, Luo K, Yu Y, Yu XF. Influence of primate lentiviral Vif and 5. Stewart S, Jablonowski H, Goebel FD,Arasteh K, Spittle M, Rios A, et al. Randomized proteasome inhibitors on human immunodeficiency virus type 1 virion pack- comparative trial of pegylated liposomal doxorubicin versus bleomycin and vin- aging of APOBEC3G. J Virol 2004;78:2072–81. cristine in the treatment of AIDS-related Kaposi's sarcoma. international pegylated 26. Schubert U, Ott DE, Chertova EN, Welker R, Tessmer U, Princiotta MF, et al. liposomal doxorubicin study group. J Clin Oncol 1998;16:683–91. Proteasome inhibition interferes with gag polyprotein processing, release, and 6. Gbabe OF, Okwundu CI, Dedicoat M, Freeman EE. Treatment of severe or maturation of HIV-1 and HIV-2. Proc Natl Acad Sci U S A 2000;97:13057–62. progressive Kaposi's sarcoma in HIV-infected adults. Cochrane Database Syst 27. Reid EG, Looney D, Maldarelli F, Noy A, Henry D, Aboulafia D, et al. Safety and Rev 2014;CD003256, issue 9. efficacy of an oncolytic viral strategy using bortezomib with ICE/R in relapsed/ 7. Koon HB, Krown SE, Lee JY, Honda K, Rapisuwon S, Wang Z, et al. Phase II trial refractory HIV-positive lymphomas. Blood Adv 2018;2:3618–26. of imatinib in AIDS-associated Kaposi's sarcoma: AIDS Malignancy Consortium 28. Hosseinipour MC, Sweet KM, Xiong J, Namarika D, Mwafongo A, Nyirenda M, Protocol 042. J Clin Oncol 2014;32:402–8. et al. Viral profiling identifies multiple subtypes of Kaposi's sarcoma. MBio 2014; 8. Little R, Wyvill K, Pluda J, Welles L, Marshall V, Figg WD, et al. Activity 5:e01633–14. of thalidomide in AIDS-related Kaposi's sarcoma. J Clin Oncol 2000;18: 29. Lin L, Lee JY, Kaplan LD, Dezube BJ, Noy A, Krown SE, et al. Effects of 2593–2602. chemotherapy in AIDS-associated non-Hodgkin's lymphoma on Kaposi's sar- 9. Polizzotto MN, Uldrick TS, Wyvill KM, Aleman K, Peer CJ, Bevans M, et al. coma herpesvirus DNA in blood. J Clin Oncol 2009;27:2496. Pomalidomide for symptomatic Kaposi's sarcoma in people with and without 30. Somsouk M, Dunham RM, Cohen M, Albright R, Abdel-Mohsen M, Liegler T, HIV infection: A Phase I/II Study. J Clin Oncol 2016;34:4125–31. et al. The immunologic effects of mesalamine in treated HIV-infected individuals 10. Dittmer DP, Damania B. Kaposi sarcoma-associated herpesvirus: immunobiol- with incomplete CD4þ T cell recovery: a randomized crossover trial. PLoS One ogy, oncogenesis, and therapy. J Clin Invest 2016;126:3165–75. 2014;9:e116306. 11. Guasparri I, Keller SA, Cesarman E. KSHV vFLIP is essential for the survival of 31. Cella DF, Tulsky DS, Gray G, Sarafian B, Linn E, Bonomi A, et al. The functional infected lymphoma cells. J Exp Med 2004;199:993–1003. assessment of cancer therapy scale: development and validation of the general 12. McNamara RP, Chugh PE, Bailey A, Costantini LM, Ma Z, Bigi R, et al. measure. J Clin Oncol 1993;11:570–9. Extracellular vesicles from Kaposi sarcoma-associated herpesvirus lymphoma 32. Hu B, Zhou Q, Wu T, Zhuang L, Yi L, Cao J, et al. Efficacy and safety of induce long-term endothelial cell reprogramming. PLoS Pathog 2019;15: subcutaneous versus intravenous bortezomib in multiple myeloma: a meta- e1007536. analysis. Int J Clin Pharmacol Ther 2017;55:329–38. 13. Matta H, Chaudhary PM. The proteasome inhibitor bortezomib (PS-341) 33. Dittmer DP. Restricted Kaposi's sarcoma (KS) herpesvirus transcription in KS inhibits growth and induces apoptosis in primary effusion lymphoma cells. lesionsfrom patients on successful antiretroviral therapy. MBio 2011;2:e00138–11. Cancer Biol Ther 2005;4:77–82. 34. Ramos JC, Sin SH, Staudt MR, Roy D, Vahrson W, Dezube BJ, et al. Nuclear 14. Boccadoro M, Morgan G, Cavenagh J. Preclinical evaluation of the proteasome factor kappa B pathway associated biomarkers in AIDS defining malignancies. inhibitor bortezomib in cancer therapy. Cancer Cell Int 2005;5:18. Int J Cancer 2012;130:2728–33. 15. Robak T,Huang H,Jin J, ZhuJ,Liu T,Samoilova O,etal. Bortezomib-based therapy 35. Host KM, Jacobs SR, West JA, Zhang Z, Costantini LM, Stopford CM, et al. for newly diagnosed mantle-cell lymphoma. N Engl J Med 2015;372:944–53. Kaposi's sarcoma-associated herpesvirus increases PD-L1 and proinflammatory 16. O'Connor OA, Wright J, Moskowitz C, Muzzy J, MacGregor-Cortelli B, cytokine expression in human monocytes. MBio 2017;8:pii:e00917-17. Stubblefield M, et al. Phase II clinical experience with the novel proteasome 36. Lechowicz M, Dittmer DP, Lee JY, Krown SE, Wachsman W, Aboulafia D, et al. inhibitor bortezomib in patients with indolent non-Hodgkin's lymphoma Molecular and clinical assessment in the treatment of AIDS Kaposi sarcoma with and mantle cell lymphoma. J Clin Oncol 2005;23:676–84. valproic Acid. Clin Infect Dis 2009;49:1946–9. 17. Suares A, Mori Sequeiros Garcia M, Paz C, Gonzalez-Pardo V. Antiproliferative 37. Casper C, Krantz EM, Corey L, Kuntz SR, Wang J, Selke S, et al. Valganciclovir for effects of bortezomib in endothelial cells transformed by viral G protein-coupled suppression of human herpesvirus-8 replication: a randomized, double-blind, receptor associated to Kaposi's sarcoma. Cell Signal 2017;32:124–32. placebo-controlled, crossover trial. J Infect Dis 2008;198:23–30. 18. Bhatt S, Ashlock BM, Toomey NL, Diaz LA, Mesri EA, Lossos IS, et al. Efficacious 38. Kedes DH, Ganem D. Sensitivity of Kaposi's sarcoma-associated herpesvirus proteasome/HDAC inhibitor combination therapy for primary effusion lym- replication to antiviral drugs. Implications for potential therapy. J Clin Invest phoma. J Clin Invest 2013;123:2616–28. 1997;99:2082–6.

AACRJournals.org Clin Cancer Res; 26(3) February 1, 2020 565

Pilot Trial AMC-063: Safety and Efficacy of Bortezomib in AIDS-associated Kaposi Sarcoma

Erin G. Reid, Adrienne Suazo, Shelly Y. Lensing, et al.

Clin Cancer Res 2020;26:558-565. Published OnlineFirst October 17, 2019.

Updated version Access the most recent version of this article at: doi:10.1158/1078-0432.CCR-19-1044

Supplementary Access the most recent supplemental material at: Material http://clincancerres.aacrjournals.org/content/suppl/2019/10/17/1078-0432.CCR-19-1044.DC1

Cited articles This article cites 37 articles, 14 of which you can access for free at: http://clincancerres.aacrjournals.org/content/26/3/558.full#ref-list-1

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at Subscriptions [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/26/3/558. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.