The Role of the Pharmacist in Optimizing Cancer Immunotherapy: A Retrospective Study of Nivolumab Adverse Events

Abstract
Background: Immune checkpoint inhibitors (ICIs) are an emerging treatment in cancer therapy for prolonging life, minimizing symptoms, and selectively targeting cancer. Program death 1 (PD-1) inhibitors, such as nivolumab, fall within this class, enabling the patient’s immune system to detect and destroy cancer. The introduction of ICIs is changing cancer therapy, with new drugs and new toxicities—an evolving area encountered by pharmacists. Objective: This study aims to compare the pattern of nivolumab-induced adverse events observed in practice, when compared with clinical trial and literature data. The secondary aim of the study is to identify the presentation and treatment modalities initiated in practice. Methods: We performed a retro- spective case note review across 2 South Australian hospitals to identify the common toxicities and symptomatic treatments experienced by patients receiving nivolumab. Results were compared with clinical trial data from product innovator Bristol-Myer Squib and other published literature. Results: Seventy patients were included in the study; of these, 60 (86%) experienced any grade adverse event(s). A total of 59 (84%) of 70 experienced mild to moderate grade 1 to grade 2 adverse events and 10 (14%) of 70 patients experienced severe grade 3 to grade 4 adverse events, displaying some consistencies with clinical trial and published literature data. Together, the prevalence of adverse events with details on presentation and treatments illustrates possible pharmacy practice strategies and areas for intervention. Conclusions: The listed prevalence of adverse events and practice strategies identified throughout this study highlights how pharmacists may assist in the identification of predictable ICI toxicities
associated with gastrointestinal, endocrine, dermatological toxicities, and fatigue.

Introduction
The emergence of cancer immunotherapy is sweeping the world, with the introduction of immune checkpoint inhibitors (ICIs), chimeric antigen receptor therapy, therapeutic vaccina- tion, and bispecific T-cell engager therapy.1 Arising with the appearance of ICIs are program death 1 (PD-1) inhibitors pem- brolizumab and increasingly popular nivolumab.1,2 These agents target the PD-1 and program death ligand 1 (PD-L1) receptor pathway.2-4The PD-1 and PD-L1 pathway introduces a new approach in cancer treatment, with the efficacy relying on the patient’s immune system to detect and destroy cancer. PD-1 inhibition reduces the immunosuppressive response when costimulated with the PD-L1 receptor expressed and often upregulated on patient cancer cells.5-7 By inhibiting the PD-1/PD-L1 receptors, T cells can recognize the patient’s cancer, leading to immune-mediated destruction of the cell.7,8ICIs provide long-term treatment opportunities, with patients reporting greater tolerability of ICIs when compared with traditional cytotoxic therapies.3,4 Traditional cytotoxic drugs can affect any cell in the body and predominantly affect tissue with high cellular turnover, including the gastrointestinal tract, bone marrow, hair, and skin. These agents can cause severe toxicities, often so distressing to the patient they areunable to continue and complete treatment.9,10 In contrast with ICIs, targeted medicines—including the commonly employed tyrosine kinase inhibitors (TKIs)—exert their mode of action by targeting specific cellular proteins vital for cancer survival and progression. While also considerably more selective and tolerable than cytotoxic therapies, each toxicity is often asso- ciated with the particular drug used.11 TKI adverse events arise from inhibiting specific “on-target” kinases, binding to nonin- tended “off-target” proteins, and also from toxic/harmful drug metabolites.11-13 In practice, patients often commence ICI ther- apy upon disease progression or discontinuation due to adverse effects with cytotoxic and/or targeted therapy.Years of experience with cytotoxic therapy allow health professionals to predict the typical gastrointestinal, hematolo- gical, and infective complications of therapy.

Compara- tively, the introduction of ICIs has revealed an emerging range of autoimmune-associated adverse events.15-20 Under normal physiological conditions, PD-1 supresses autoimmunity; hence, inhibiting this process can generate autoimmune- associated adverse events—often predictable in presentation.21 The increasing use of anti-PD1 therapy highlights the impor- tance of wide-spread understanding and therefore education of health professionals on the adverse events of these medicines and how they differ from those associated with cytotoxic therapy.15,22,23The growing evidence of immune-related toxicities has researchers encouraging health professionals to closely moni- tor patients using ICIs to enable further understanding of the safety profiles of these medicines.21,22 Given the pharmacists accessibility in primary and acute care, pharmacists are suita- bly placed to assist in therapy optimization initiatives. With skills in interpreting laboratory biochemistry, monitoring toxi- cities, and providing counseling regarding medication-related safety concerns, pharmacists could assist in identifying and reducing the severity of immune-related adverse events in patients taking ICIs. However, the lack of awareness and lim- ited clinical guidelines for pharmacists working in this field constitutes a significant barrier that often prevents the delivery of these clinical services.The primary aim of this study was to capture the adverse events of nivolumab—one of the most commonly used ICIs— across 2 South Australian public hospitals. This study assesses the prevalence, type, and severity of adverse events experi- enced. A secondary aim of the study was to identify how nivo- lumab induced adverse events present in practice and identify common treatments applied by the attending medical teams. As a result, the outcomes of this study will serve as an educational resource for pharmacists by providing details on adverse event prevalence and possible pharmacists-led practice points.The study was conducted at 2 South Australian hospitals, the Royal Adelaide Hospital (RAH) and Flinders Medical Centre (FMC). Study protocol and data collection template were cre- ated using Microsoft® Word 2017 (version 15.26). Patientsreceiving nivolumab for the retrospective study were identified from electronic dispensing records in iPharmacy software (ver- sion 8.1.0) across the RAH and FMC between March 2015 and 2018.

Full medical records of patients were then assessed against inclusion and exclusion criteria to determine eligibility in the study. Inclusion criteria required the patient to receive at least 1 cycle of nivolumab in accordance with an established local cancer treatment protocol. Patients were excluded if nivo- lumab was dispensed but not administered, treatment occurred outside of the study time frame, or medical record data were insufficient to complete the data collection form. Patients enrolled in clinical trials were also excluded from the study to capture a real-world incidence of adverse events reflective of standard nivolumab therapy in established cancer protocols. To ensure consistency of data collection, standardized data collection forms were utilized by 1 project investigator across both RAH and FMC. This enabled specific collection of patient demographics including cancer type, past medical history, age, and gender; adverse events experienced, severity and fre- quency; and routine biochemistry to confirm adverse events. Where possible, the adverse events were graded in accordance with Cancer Terminology Criteria for Adverse Events as con-sistent with the nivolumab (Opdivo®) product information.24,25Captured patient biochemistry for the study included random cortisol, alanine transferase, aspartyl transferase, thyroid- stimulating hormone (TSH), free levothyroxine, creatinine, and albumin. Patient adverse event data that contained abnormal- ities in baseline biochemistry and events in those with a prior diagnosis of thyroid, liver, and/or kidney disease were excluded to minimize potential for false-positive results, with adverse events deemed likely related to the patients’ underlying condition, rather than nivolumab treatment. Where possible, details of medication and pharmacotherapy utilized in treat- ment for adverse events were obtained throughout the data collection process.The following sources enabled data collection for the study: patient clinical notes containing inpatient admission notes, oncologist reviews, and outpatient visit records; Open Archi- tecture Clinical Information System, clinical manager version7.1.0.106 to obtain biochemistry, patient handover notes, and clinical summaries; and iPharmacy records to determine nivo- lumab dose, frequency, and route of administration.

To assess for the total incidence of adverse events across all cancer types, the data were pooled together to give a total population pre- valence for the study and experienced adverse events. Data were collated and analyzed using Microsoft Excel (version 15.26), stratifying by patient cancer types, experienced adverse events, adverse event severity, and patient characteristics— age, sex, and biochemistry.Details of patient adverse event presentation, diagnosis, and treatments were collated with patient demographics. Individual case summaries are presented to provide real-world examples of patient adverse events, presentations, and treatments. The cases were selected from those that included detailed documen- tation regarding adverse effect management to demonstrate a spectrum of events and treatments typical of the broader patientcohort. Possible pharmacist-led interventions were identified by the authors based upon the findings of the retrospective review and confirmed with data ascertained from medical guidelines and other available literature.Quantitative descriptive statistics were used to characterize cohort demographics, cancer, and treatment characteristics. Adverse events identified in our patient cohort, including total prevalence and adverse event severity (grade 1-4), were com- pared with data presented by innovator Bristol-Myer Squib in the nivolumab product information and media press release.25,26 Data were also compared with a pooled analyses of nivolumab adverse events reported in the literature.

Results
iPharmacy revealed 86 patients dispensed nivolumab between March 2015 and 2018; 16 patients were excluded due to insuf- ficient data or failure to proceed with ICI treatment (Figure 1).The median age was 63.4 years (range: 17-85 years), with the majority of patients being males (64%; Table 1). Patients undertook a median 15 cycles (range: 1-60) of nivolumab treat- ment. All patients received nivolumab dosed at 3 mg/kg by intravenous (IV) infusion once per fortnight (14 days). Twenty-six (37%) patients remained active in treatment at the time of the study, 36 (51%) patients had discontinued due to disease progression, and 8 (11%) patients discontinued due to adverse events.Adverse events of any grade occurred in 60 (86%) of 70 patients, with 59 (84%) of 70 patients experiencing minor to moderate grade 1 to grade 2 adverse events, resulting in the need for simple intervention, initiation of corticosteroids, mon- itoring for symptom progression, and/or delaying subsequent nivolumab dosing; 10 (14%) of 70 patients in the study had severe grade 3 to grade 4 adverse events resulting in holding treatment, treatment cessation, initiation of corticosteroids, and/or additional medical interventions. Identification ofsevere grade 3 toxicities prompted discontinuation of treatment in 8 (11%) of 70 patients. In our cohort, we identified 3 patients who experienced hepatitis: 2 had experienced pneumonitis and single cases of nephritis, central nervous system derangement (delirium and hallucination), and colitis. Grade 3 toxicity occurred after a mean of 8.3 cycles; 1 patient was transferred to another hospital and lost to follow-up and 6 patients required inpatient treatment and discontinuation of nivolumab therapy. One patient who had grade 3 toxicity died shortly after nivo- lumab cessation.When comparing with literature and clinical trials data, the results showed a numerically higher prevalence of fatigue, nausea, rash, adrenal insufficiency, and pneumonitis (Table 2).

The largest difference was the increased prevalence of fatigue, resulting in an overall 21% increase. An increase of 6% was recorded in patients having adrenal insufficiency; however, only 1 patient received confirmed diagnosis of hypophysitis due to nivolumab.The most common adverse event was fatigue, experienced by 36 (51%) of 70 patients. The most common adverse event requiring medical intervention was gastrointestinal toxicity, experienced by 34 (49%) of 70 patients; 22 (31%) of 70 patients experienced nausea, a further 8 reported at least 1 episode of vomiting, with 2 patients requiring hospitalization for colitis. Treatment of gastrointestinal toxicity varied on pre- sentation and severity of symptoms. Hypothyroidism was documented in 6 (9%) of 70 patients after median time of 13cycles (range: 2-39); 2 (3%) of 70 patients experienced transi- ent hyperthyroidism/thyrotoxicosis, shortly after developing into hypothyroidism, with no cases requiring pharmacotherapy. Most adverse events observed were transient and required only monitoring prior to subsequent dosing of nivolumab and/events in the treatment population, outlining the clinical pre-sentation and treatment employed. The observed adverse events and practice points are likely generalizable to other anti-PD-1 ICIs, including pembrolizumab. The symptoms, treatments, and laboratory abnormalities listed in Table 3 provide insight into areas for pharmacist review; identifying these abnormalities and symptoms early may assist in pre- venting patient harm from nivolumab treatment. The derma- tological, endocrine, hepatological, and gastrointestinal systems are key adverse event-prone systems that could be monitored using laboratory testing and enquiring about patient symptoms. Pharmacists could utilize this information and contribute by identifying, preventing, and referring patients with the toxicities discussed.

Discussion
This study identified the common adverse events associated with nivolumab treatment in practice, documenting a numeri- cally higher prevalence of nivolumab-associated fatigue, nau- sea, rash, adrenal insufficiency, and pneumonitis, compared with the literature. Clinical trials data from the manufacturers product information and published literature were considered suitable comparators for the project, as all trial patients were dosed at 3 mg/kg every 14 days and experienced cancer types similar to the study population.25,27It is important, however, to interpret these findings in the context of the limitations of this study, which can be drawn from the retrospective design and limited sample size of the study cohort. Data comparators utilized in this study arise from prospective trials. As a result, the differences in study design have potential to introduce selection and recording bias in data collection. Furthermore, the retrospective design of the study placed a limitation on the ability to accurately define each adverse event recorded and treatment modality employed. Often the adverse event severity was not recorded or defined by the treating medical team, and subsequent details on adverse event treatment (dose, duration, and/or tapering) were not described. The small sample size of the study population lim- ited the ability to perform statistical calculations and conclude whether significant differences can be seen between the study population and comparator groups.The differing clinical presentations and varying degree of severity of adverse events experienced by patients taking nivolumab identified in this study highlight the need for phar- macists to be aware of how nivolumab-related adverse events may present and the associated recommended monitoring. This study describes methods of identifying adverse events associated with the gastrointestinal, dermatological, endo- crine, and hepatological systems and outlines the common treatments for adverse events used in practice.

Consistent with the literature, this study demonstrates topical and oral corti- costeroids remain one of the most commonly used pharmaco- logical interventions to treat anti-PD-1 adverse events.23,28 The intention of using corticosteroids is to supress the immune response accountable for the adverse events inflicted on the patients’ noncancerous tissue.Many patients experience nausea during immune checkpoint therapy, with symptomatic treatment typically resolving symp- toms successfully.28,30,31 Our study identified many different antiemetics used in patients with little to no superior agent identified in published literature and practice. All patients treated with dopamine antagonists (metoclopramide and pro- chlorperazine) when required displayed successful symptom resolution. Lower gastrointestinal toxicity, defined as diarrhea and/or colitis, is due to underlying inflammation of the colon, most frequently presented as mild diarrhea; however, escala- tion to severe colitis is possible.30 All patients presenting with mild diarrhea were successfully treated with loperamide, as guided in published literature.30,28 While many cases of nivolumab-induced gastrointestinal toxicity are mild, pro- longed, severe gastrointestinal adverse events and those unre- sponsive to simple intervention require urgent medicalreferral.30,28 In all cases, the treating clinician should consider underlying bacterial/viral causes and undertake suitable diag- nostic procedures.30,28 Treatment escalation typically involves delaying subsequent nivolumab dosing, initiating oral or IV corticosteroids, and if unresponsive, escalating to immune modulator therapy such as infliximab.23,30,28We identified 6 patients experiencing hypothyroidism using routine monitoring of weight, biochemistry (TSH and free thyr- oxine), fatigue, and appetite—key indicators of hypothyroid- ism.29,32,33 The results demonstrated consistency to the literature, with hypothyroidism or transient hyperthyroidism– hypothyroidism occurring between 2 weeks and 19 months.18 As such, monitoring for these symptoms and patient biochem- istry could assist in early detection of hypothyroidism and potentially minimize any further complication to nivolumab therapy.29 All cases of hypothyroidism received substitution treatment with levothyroxine with current guidelines primarily focusing on minimizing the symptoms of the hypothyroid- ism.

The study captured 5 cases of adrenal insufficiency; however, 4 of the reported cases occurred in patients taking concurrent glucocorticoids (dexamethasone or prednisolone at varying doses) to prevent chronic obstructive pulmonary dis- ease exacerbations and/or treat immune-related adverse events, known contributors to supress patient cortisol.35,36 The diag- nosis of adrenal insufficiency occurred following routine cor- tisol testing in 2 of 5 patients and after presentation of symptoms including hypotension, hyponatremia, and confu- sion consistent with Addisonian crisis in 3 of 5 patients. Four of 5 patients received oral hydrocortisone dosed in morning and afternoon; no patients with endocrine toxicities in this study discontinued treatment.cycle 3 from 49 units/L to 68 units/L and then back to baseline. Transient rise in AST in cycle 51 from 26 U/L to 128 U/L and then back to baseline.The study identified 19 (27%) of 70 patients with rash (any grade), a common adverse event associated to ICI therapy. Cur- rent guidelines recommend using topical corticosteroids and/or emollient creams, with many patients in our study receiving topical corticosteroids and monitored over time.28,37 One patient reported worsening symptoms after prescribed betamethasone diproprionate and was successfully treated by escalating therapy to combination of betamethasone and calcitriol (Daivobet®).Pruritus was the second most common dermatological toxicity experienced in 10 (14%) of 70 patients.

Treatment comprised of antihistamines, topical corticosteroids, and topical emollients as consistent with the literature.28,37 For all patients, symptoms resolved successfully with no patient discontinuing due to der- matological toxicity. Grade 3 and grade 4 toxicities can occur in practice and require immediate referral to dermatology specialty, with most cases resulting in delayed subsequent dosing or ceas- ing the ICI and requiring systemic corticosteroid therapy.16,37,38Studies suggest fatigue is one of the mostly commonly expe- rienced adverse events identified as a result of anti-PD1 ther- apy.27,39,40 Despite this, there are currently no effective treatment modalities established or utilized in practice. Clinical practice suggests that when patients present with severe cases or worsening cases of fatigue, assessment for hypothyroidism and laboratory abnormalities in electrolytes should be under- taken.21,41 Many patients throughout the study reported chang- ing diet or increasing duration of rest to overcome fatigue when experienced. No patients were discontinued from treatment due to the symptoms of fatigue.The results show transient rises in patient transferase concentra- tion consistent with literature data.25,27 This highlights the impor- tance of understanding clinical context—liver metastasis and other drugs can also increase transaminase and should be accounted for.42,43 Transient rises in transaminases may have a low risk of patient harm; however, the results illustrate a strong importance of patient monitoring with 3 cases of asymptomatic hepatotoxicity prompting treatment discontinuation identified in our study.44 The results provide an important reminder to com- pare patient biochemistry with clinical circumstance and patient baseline, as minor elevations may cause no harm or symptoms.Guidelines suggest undertaking periodic blood tests when treated with nivolumab, a robust method for preventing the development of severe toxicities in current practice.25,45 Treatment of patients with hepatotoxicity typically involves monitoring transaminase level and holding doses of the ICI. Cases of >grade 3 hepatotoxi- city result in treatment cessation and treatment with systemic corticosteroids.45 The study revealed 3 patients having hepato- toxicity with all cases discontinuing treatment and a single patient undertaking short-term systemic glucocorticoid therapy.

Conclusions and Recommendations
The results show a numerically higher prevalence of nivolumab-associated fatigue, nausea, rash, adrenal insuffi- ciency, and pneumonitis when compared with clinical trial data and other published literature. The study identified opportuni- ties for pharmacist intervention in relation to gastrointestinal, endocrine, hepatological, and dermatological toxicities. Currently, limited pharmacist adverse event awareness (pre- valence, presentation, and treatment modalities) represents a barrier to pharmacist intervention in ICI therapy. As illustrated by the strategies presented throughout this article, ICIs present a great opportunity for pharmacists to assist in mitigating adverse effects and optimizing the outcomes of cancer treat- ment. We encourage pharmacists to support patients receiving nivolumab therapy in managing their toxicities and refer patients to their treating medical team if symptoms are progres- sively worsening, with intention to prevent to moderate–severe toxicities (grades 2-4) from occurring.
We identify the expanding use of ICIs in practice to provide the pharmacist AUNP-12 frequent opportunity to contribute toward multi- disciplinary care and recommend all pharmacists support patients undertaking anti-PD-1 therapy by assessing for and managing the adverse events discussed throughout this study.