oe VOL. 15, NO. 3, august 2016 25

landmarks Reports of recent research

Report from the Canadian Association of Medical Oncology Annual Scientific Meeting

side effectsthe rOLe Of ObserVatiONaL studies iN medicaL ONcOLOgy research

Katarzyna Jerzak and Kathleen I. Pritchard, sunnybrook Odette cancer centre, toronto, ON, canada

TrIal Summary: Febrile neutropenia with FEC-TVeitch Z, bashir a, tilley d, et al. a retrospective provincial analysis of real-world outcomes and febrile neutropenia rates in women using adjuvant fec-taxane (fec-t) for node-positive, her2-negative breast cancer in community and tertiary cancer centers. canadian association of medical Oncology annual scientific meeting, 2016, abstract 16.

This study compared rates of febrile neutropenia (FN), overall survival (OS), disease-free survival (DFS), and hor-mone adherence rates (HA) in women under age 50 and women over 50 years old undergoing FEC-T (5-flurouracil+epirubicin+cyclophosphamide followed by adjuvant che-motherapy) for stage II/III, node-positive, HER2-negative breast cancer. A total of 919 women (274 under age 50 and 645 over age 50) treated with FEC-T in community and tertiary Alberta cancer centres between 2008 and 2012 were included, 88.2% (811) of whom completed recom-mended treatment.

Results: Rates of FN were similar between patients under age 50 (15.7%; 95% CI: 11.4, 20.0%) and 50 or over

(19.4%; 95% CI: 16.3, 22.5%; p=0.202) and occurred with FEC (cycle 1–3; 51%) and taxane (cycle 4–6; 49%) regi-mens equally. DFS was higher in the under-50 cohort (90.9%) than in patients 50 and over (83.8%), although this was nonsignificant (p=0.251). Differences in breast-cancer-specific OS showed a trend to significance (95% for women under 50 vs 88.1% for women 50 and over; p=0.124). OS was similar in community (87.3%) and tertiary (91.3%) set-tings (p=0.137). Hormonal adherence rates were signifi-cantly higher (p<0.001) for tamoxifen to aromatase inhibi-tor (AI) switch compared to monotherapy at yearly time-points. Two treatment-related deaths occurred in the under-50 age group (leukemic and treatment related). In this cohort, FN rates occurred with similar frequency in the FEC and taxane groups, regardless of age or place of treat-ment. Breast cancer-specific OS and DFS were similar for women under 50 and women 50 or older. Hormone adher-ence rates were significantly higher for those on a hormonal switch strategy compared to monotherapy.

26 oe VOL. 15, NO. 3, august 2016

landmarks

TrIal Summary: Carcinoid heart diseasemichael J and rayson d. carcinoid heart disease: clinical and patho-logic characteristics. canadian association of medical Oncology annual scientific meeting, 2016, abstract 24.

Metastatic, functional neuroendocrine tumours (NETs) release bioactive substances into the circulation. Carcinoid syndrome arises due to high levels of circulating serotonin. Carcinoid heart disease (CHD) occurs as a result of long-standing carcinoid syndrome, manifesting as confluent, plaque-like lesions on valvular endocardium, increasing the risk of symptomatic valvular insufficiency and cardiac failure. This study describes the clinical and pathologic characteris-tics and outcomes of a consecutive series of patients with CHD treated at the Queen Elizabeth II Health Sciences Centre in Halifax, Nova Scotia. A chart review of the 11

patients diagnosed with CHD between January 1, 2000, and July 1, 2015, was performed alongside descriptive anal-ysis of pathologic, clinical, and biochemical variables.

Results: Median OS for the entire cohort from time of NET diagnosis was 69 months (range 9 to 217 months). Four patients were diagnosed with CHD within 3 months of initial NET diagnosis, and all underwent valvular surgery within 1 year, with a median postoperative survival of 7.5 months (range 3 to 54 months). The remaining 7 patients had a median time to development of CHD of 57 months (range 18 to 139 months). Of these, 2 were unfit for surgery, 1 died of an unrelated illness and 1 is being observed. Three remaining patients underwent cardiac surgery and had a median postoperative survival time of 30 months (range 12 to 66 months). Median time to hospital discharge was 5 days (range 4 to 8 months) with a 0% 30-day mortality rate.

TrIal Summary: Choice of endpoints to assess nausea and vomitingNg t, mazzarello s, Wang Z, et al. impact of study endpoint selec-tion on the results of breast cancer trials evaluating chemotherapy-induced nausea and vomiting. canadian association of medical Oncology annual scientific meeting, 2016, abstract 26.

Concerned that the choice of study endpoint could poten-tially change the interpretation of clinical trial results, this study evaluated chemotherapy-induced nausea and vomiting (CINV) rates using different endpoints on a single dataset from a prospective cohort. It also explored the frequency with which published pharmaceutical and non-pharmaceu-tical company-funded CINV studies involving breast cancer patients included measures of nausea in their primary study endpoint. Data from 177 breast cancer patients receiving anthracycline and cyclophosphamide-based chemotherapy was used to estimate CINV control rates using the 15 most commonly reported CINV endpoints. In a sample of 30

CINV trials involving breast cancer patients that met the final eligibility criteria in a recently published systematic review, the authors examined the number of trials that used a primary endpoint that measured nausea control.

Results: CINV control rates ranged from 12.5% (95% CI 7.6%–17.4%) for total control (no vomiting, no nausea and no rescue medication use) to 77.4% (95% CI 71.2%–83.6%) for no vomiting in the overall period. Similar differ-ences were found in the acute and delayed periods. Primary study endpoint(s) measuring nausea control were more commonly found in trials that were not funded by pharma-ceutical companies (9/18, 50%) compared to those funded by pharmaceutical companies (1/12, 8.3%). The choice of trial endpoint has important effects on reported CINV control rates and could significantly impact interpretation of results. Primary endpoints of studies, including those mandated by regulatory bodies, should reflect patient expe-rience with nausea. More comprehensive reporting of end-points would allow for important cross-trial comparisons.

TrIal Summary: acute pain syndrome with taxanesfernandes r, mazzarello s, mohammed s, et al. a multicentre study to investigate the natural history of taxane acute pain syndrome (taPs) in patients receiving taxane-based chemotherapy for breast or prostate cancer . canadian association of medical Oncology annual scientific meeting, 2016, abstract 28.

Taxane acute pain syndrome (TAPS) is characterized by myalgia and arthralgia starting 24–48 hours after receiving taxane-based chemotherapy and lasting up to 7 days. Despite its negative impact on patient quality of life (QoL), its characteristics and natural history remain poorly defined. This study evaluated the persistence, severity and impact of TAPS on health related QoL. Eligible patients with breast or prostate cancers commencing taxane-based chemotherapy completed the Functional Assess-ment of Cancer Therapy-Taxane (FACT-T) and Brief Pain Inventory (BPI) questionnaires, and kept a pain medication diary every day for 1 week after each chemotherapy infusion. TAPS

was defined as myalgias and arthralgias on these questionnaires.Results: From March to December 2015, 25 of the 52

patients enrolled (21 with breast cancer and 4 with prostate cancer) completed the study. TAPS was reported in 66% of breast cancer patients, starting 24 to 72 hours (range 24 to 96 hours) after treatment infusion, and reaching a peak by day 3 (range 1 to 5 days). TAPS was reported more often in the legs and back. Table 1 shows the characteristics and clinical features. There were no dose reductions, delays or treatment discontinuations required due to TAPS. Medica-tions used to treat TAPS included opioids in 5 patients and nonsteroidal antiinflammatory drugs (NSAIDs) in 9. Mean change in BPI “worst pain” score was +1.61 (p=0.014) and on FACT-T was -6.9 (p=0.017) from baseline to final infusion cycle. TAPS is a common toxicity and is associated with a negative impact on QoL. Further data will help define pre-disposing risk factors. Prospective patient-reported outcome assessments are crucial to help individualize treatment strat-egies, as well as improve management of TAPS.

oe VOL. 15, NO. 3, august 2016 27

In BrIEFalready known• randomized controlled trial design is not suited to

all questions in oncology.

What these studies showed• retrospective, prospective and cross-sectional obser-

vational studies provided important information about side effects of chemotherapy, including neu-tropenia, nausea and vomiting, taxane-induced pain syndrome, as well as the rare condition of carcinoid heart disease.

Next steps• results of observational studies should inform future

clinical trial design and guide treatment selection for particular subgroups of patients.

landmarks

CommENTary: Data from large randomized controlled trials (RCTs) represent the highest quality evidence when establishing new therapies in oncology.1,2 Unfortunately, it is not always possible to conduct a large randomized study; high costs, low recruitment rates and long time periods of evaluation are just a few of the limitations. Further, RCT trial design is not suited to all research questions.

The importance of another method, observational research, is well illustrated by 4 abstracts that were present-ed in the 2016 Canadian Association of Medical Oncology (CAMO) Annual Meeting. In 1 of these abstracts, Michael et al conducted a small retrospective chart review to describe the natural history of carcinoid heart disease in 11 patients who were diagnosed with this rare condition over a 15-year time period.3 Investigators made a meaningful con-tribution to the literature by identifying minimal periopera-tive mortality among 6 patients who required valvular sur-gery.3 Clearly, a RCT would not be suitable to evaluate the role of cardiac surgery in this rare condition.

Two abstracts regarding side effects of chemotherapy highlight the importance of prospective observational stud-ies in oncology. Using a cohort of 52 patients with either breast or prostate cancer, Fernandes et al. describe the fea-tures of TAPS and its impact on QoL.4 Such prospective data is critical, not only to define the natural history of this poorly understood syndrome, but also to inform the design of future RCTs that compare various therapeutic strategies. Currently, the optimal treatment for TAPS remains unknown.

In another prospective cohort study, Ng et al. describe the incidence of CINV in 177 patients receiving either cyclophosphamide or anthracycline-based therapy.5 Investi-gators rated “CINV control” using 15 definitions that are commonly employed in clinical trials. Depending on the definition used, between 12.5% to 77.4% of patients were reported to have “control” of their nausea and vomiting.5 Through this study, investigators highlight a need to unify the definition of patient-reported outcomes in clinical trials. This is critical to better inform patients, physicians and reg-ulatory bodies regarding the toxicity of both traditional and newer targeted systemic therapies.6,7

In a fourth abstract, a cross-sectional study design was

Table 1. Taxane acute pain syndrome (TaPS) characteristics and clinical factors

TC (n=10) FeC-D (n=2) Single-agent paclitaxel (n=9) aC-P or aC-D (n=3)

Treatment setting adjuvant adjuvant metastatic adjuvant

TAPS incidence 80% 50% 22% 100%

TAPS duration (days) 1–5 (median: 4; mean: 3.26)

2–4 (median: 3; mean: 3)

2–5 (median: 3; mean: 3.4)

3–5 (median: 4; mean: 3.72)

Growth factor use (n) 7 1 0 1

Taxane dosing at 100 mg/m2 no yes no yes

Oral steroids use (n) 8 1 0 2

TC=cyclophosphamide plus docetaxel; FEC-D=5-fluorouracil+epirubicin+cyclophosphamide followed by docetaxel; AC-D=doxorubicin+cyclophosphamide followed by docetaxel; AC-T=doxorubicin+cyclophosphamide followed by paclitaxel

employed to determine rates of febrile neutropenia among women receiving adjuvant FEC-taxane (FEC-T) for early breast cancer in Alberta.8 Given that patients with medical comorbidities and those at advanced age are often under-represented in RCTs, treatment efficacy may be overesti-mated and toxicity may be underappreciated.9,10 Hence, this province-wide study highlights the power of large databases in exploring the toxicity of cancer therapies in a broad patient population and under real-world conditions.

All 4 featured abstracts used observational research methods to tackle important clinical questions that are ill-suited to RCT design. Given an undisputed role for both retrospective and prospective databases in oncology research, efforts should be made to unify measures of QoL patient-reported outcomes, and treatment efficacy. By cre-ating an infrastructure for data sharing among institutions and provinces across Canada, we can accelerate progress toward answering key questions in a rapidly changing and diverse field of oncology.

28 oe VOL. 15, NO. 3, august 2016

landmarks

references1. Booth CM, Cescon DW, Wang L, et al. Evolution of the Randomized Controlled

Trial in Oncology Over Three Decades. J Clin Oncol. 2008; 26(33):5458-5464.2. Guyatt G, Rennie D, Meade M and Cook D. Users’ guides to the medical literature.

American Medical Association, 2015.3. Michael J, Rayson D. Carcinoid heart disease: clinical-pathologic characteristics.

Canadian Association of Medical Oncology Annual Scientific Meeting, 2016, Abstract 24.

4. Fernandes R, Mazzarello S, Ibrahim MFK, et al. A multicentre study to investi-gate the natural history of taxane acute pain syndrome (TAPS) in patients receiving taxane-based chemotherapy for breast or prostate cancer. Canadian Association of Medical Oncology Annual Scientific Meeting, 2016, Abstract 28.

5. Ng T, Mazzarello S, Wang Z, et al. Impact of study endpoint selection on the results of breast cancer trials evaluating chemotherapy-induced nausea and vomiting. Canadian Association of Medical Oncology Annual Scientific Meeting, 2016, Abstract 26.

6. Wehrlen L, Krumlauf M, Ness E, et al. Systematic collection of patient reported outcome research data: A checklist for clinical research professionals. Con-temp Clin Trials. 2016;48:21-9.

7. Gnanasakthy A, DeMuro C, Clark M, et al. Patient-Reported Outcomes Labeling for Products Approved by the Office of Hematology and Oncology Products of the US Food and Drug Administration (2010-2014). J Clin Oncol. 2016;34(16):1928-34.

8. Veitch Z, Bashir A, Tilley D, et al. A retrospective provincial analysis of real-world outcomes and febrile neutropenia rates in women using adjuvant FEC-taxane (FEC-T) for node positive, HER-2 negative breast cancer in community and tertiary cancer centers. Canadian Association of Medical Oncology Annual Scientific Meeting, 2016, Abstract 16.

9. Giordano SH. Comparative effectiveness research in cancer with observational data. Am Soc Clin Oncol Educ Book. 2015:e330-5.

10. Hershman DL, Wright JD. Comparative effectiveness research in oncology methodology: observational data. J Clin Oncol. 2012;30(34):4215-22.