Background

1. Randomized Controlled Trials Examining Continuous Versus Intermittent Strategies Of Delivering Systemic Treatment For Untreated Metastatic Colorectal Cancer: A Meta-Analysis From The Cancer Care Ontario Program In Evidence-Based Care Scott R. Berry1, Roxanne Cosby2, Timothy R. Asmis3, Kelvin K. Chan1, Nazik Hammad4, Monika K. Krzyzanowska5 1. Sunnybrook Odette Cancer Centre, University of Toronto, Toronto, ON; 2. Program Evidenced Based Care, McMaster University, Hamilton, ON; 3. The Ottawa Hospital, Ottawa, ON; 4. Cancer Center of Southeastern Ontario, Kingston, ON; Princess Margaret Cancer Centre, University of Toronto, Toronto, ON

2. Background

3. Background • A number of randomized trials have now studied intermittent strategies of administering first line systemic therapies to patients with unresecatble metastatic colorectal cancer (mCRC) to ameliorate the toxicities that patients experience. • These intermittent chemotherapy strategies involve: • an induction period with chemotherapy (with or without a biologic) • followed by a period where one or all of the chemotherapy drugs are discontinued, • followed by re-introduction of the induction chemotherapy at some point.

4. Background • Important goal of these “stop-and-go” strategies is to reduce side effects and improve patients’ quality of life • ? impact of these strategies on efficacy. • The Gastrointestinal Disease Site Group (GI DSG) of the Cancer Care Ontario (CCO) Program in Evidence Based Care (PEBC) decided that a systematic review of the evidence and a synthesis of the available data by meta-analysis would be useful in helping clinicians’ recommend appropriate treatment strategies for patients with mCRC.

5. Background • A recent meta-analysis (Pereira et al, ESMO 2012) comparing intermittent and continuous chemotherapy strategies concluded that continuous chemotherapy had a significant, albeit modest, improvement in survival (HR, 0.90; 95%CI, 0.82-0.99, p=0.03) over intermittent chemotherapy. • However, there are two main issues with this meta-analysis: • It only included five trials one of which was the COIN-B trial (Wasan et al, ASCO GI 2012). COIN-B does not have a true continuous chemotherapy arm and was therefore not included in the present systematic review. • Secondly, the forest plot of the Pereira et al. meta-analysis shows one trial (Maughan, BMJ, 2003) to have an overall survival HR of 0.87 in favour of continuous chemotherapy when, in fact, the overall survival HR in this trial is 0.87 in favour of intermittent chemotherapy. • Given that the Maughan et al trial accounts for 17.1% of the result and the confidence interval for the HR is so close to 1.00, it is possible that the conclusion of the Pereira et al. meta-anlaysis will not be maintained when this error is corrected.

6. QUESTION: What is the impact of reducing exposure to systemic therapy with intermittent administration strategies on efficacy and toxicity (including length and quality of survival) in inoperable advanced colorectal cancer?

7. Methods

8. Methods • A systematic search strategy was developed for MEDLINE, EMBASE and ASCO and ESMO proceedings was developed to identify relevant studies • Studies identified through the search were reviewed by one physician member and the methodologist of the CCO GI DSG and selected if they: • Were published English-language abstracts or fully published reports of Phase II or III Randomized Controlled Trials that comparedcontinuous chemotherapy to an intermittent strategy of chemotherapy, with or without maintenance chemotherapy, in adult patients with metastatic colorectal cancer • Included at least one of the outcomes of interest.

9. Methods • When clinically homogenous results from two or more trials were available, the data was pooled using the Review Manager software (RevMan5.2) provided by the Cochrane Collaboration • Hazard ratios (HRs), as the preferred statistic for pooling time-to-event outcomes (Parmar, Stat Med, 1998), were extracted directly from the most recently reported trial results. • The variances of the hazard ratio estimates were calculated from the reported confidence intervals (CI) using the methods described by Parmar et al. • A random effects model was used for all pooling.

10. Methods • Statistical heterogeneity was calculated using the χ2 test for heterogeneity and the I2 percentage. • A probability level for the χ2 statistic less than or equal to 10% (p≤0.10) and/or an I2 greater than 50% were considered indicative of statistical heterogeneity.

11. Methods • Meta-analyses were only conducted on the overall survival outcome. • HR < 1.0 - patients receiving intermittent chemotherapy had a lower probability of experiencing an event (death) • HR >1.0 - patients receiving continuous experienced a lower probability of an event.

12. Methods • A meta-analysis was conducted on all trials (single agent and combination chemotherapy trials) followed by an analysis of combination chemotherapy trials only • Further analyses were conducted to determine the impact of the different induction / continuous and maintenance strategies used in the different trials on Overall Survival

13. Results

14. Results • 10 trials with 4296 patients were identified that met the selection criteria • 7 trials with 4009 patients had available Overall Survival (OS) Hazard ratios (HRs) • Tables 1 and 2 outline key elements of these trials

15. Table 1: Randomized Trials of Continuous vs Intermittent Strategies of Administering Systemic Therapies for Untreated mCRC

16. Table 2: Randomized Trials of Continuous vs Intermittent Strategies of Administering Systemic Therapies for Untreated mCRC OS= Overall survival, DFS = Disease Free Survival, DDC = duration of disease control, TTF = time to treatment failure , NS = non-significant, NR = Not Reported

17. Overall Meta-Analysis Results * n for trials with available hazard ratios

18. All Trials

19. All Trials With No Maintenance Therapy

20. All Trials with Maintenance Therapy

21. Meta-analyses By Types of Therapy • Following groups considered but only those in red had more than 1 trial with an available OS HR: • Single Agent Trials • All Combination Chemotherapy Trials and Combination Chemotherapy Trials with: • 5FU Maintenance • No Maintenance • Biologic Maintenance • 5FU and Biologic Maintenance

22. Meta-Analyses By Types of Therapy * n for trials with available hazard ratios

23. Combination Chemotherapy Trials

24. Combination Chemotherapy Groups

25. Sensitivity Analyses 2 further sensitivity analyses were performed among trials with no maintenance therapy: • SENSITIVITY ANALYSIS 1 – Only Trials With No Maintenance Therapy until progression (ie excluding Labianca trial) • Among trials with no maintenance therapy, one trial (Labianca 2011) had a unique design compared to all other trials – rather than being off therapy until progression – patients on the intermittent arm had “2 months off / 2 months on” of FOLFIRI, which is not an optimal regimen for patients • SENSITIVITY ANALYSIS 2 - Only Combination Chemotherapy Trials With No Maintenance Therapy Until Progression • Sensitivity analysis 1 includes one monotherapy trial so another sensitivity analysis was done excluding it

26. Trials with No Maintenance Therapy (Sensitivity Analysis 1 – Labianca Trial Excluded)

27. Trials with No Maintenance Therapy (Sensitivity Analysis 2 – Labianca and Maughan Trials Excluded)

28. Summary – All Analyses of Trials With No Maintenance Therapy

29. Results – Summary (Efficacy) • Meta-analysis demonstrates no statistically or clinically significant survival difference between the continuous and intermittent strategies (HR, 1.02; 95%CI, 0.95-1.10, p=0.62). • No analyses based on the type of continuous therapy or maintenance therapy in the intermittent arm, demonstrated astatistically or clinically significant difference in overall survival between the two strategies

30. Results – Summary (Toxicity) • Although differences in grade 3/4 toxicities were noted between the two treatment strategies, these toxicity assessments reflect maximal levels of toxicity experienced during exposure to the treatment on that arm of the trial. • These measures are important, but for patients on intermittent treatment, duration of exposure to toxicity, or ability of patients to recover from the toxicities after induction treatment are also important and are likely better captured in quality of life assessments.

31. Results – Summary (QOL) • Of the two trials that measured quality of life, Maughan et al showed no differences between the two arms • Adams et al showed no detriment in the intermittent arm and several benefits for the intermittent chemotherapy at 24 weeks

32. Discussion • Given that the trials included in this systematic review included a variety of maintenance strategies (including no maintenance treatment) a definitive recommendation regarding an optimal maintenance strategy is not possible. • Analyses of strategies that did not use any maintenance therapies did not demonstrate any statistically or clinically significant detriment in overall survival. • This approach may be preferred by patients as it offers them a complete break from treatment.

33. Discussion All but one of the intermittent strategies had the following parameters and would be reasonable guidelines to consider when using an intermittent strategy: • 12-18 weeks of induction • Monitored with imaging at least every 8-12 weeks • Reintroduction of therapy at disease progression. Adaptation of a strategy to individual circumstances should be considered • Eg. a longer induction period or closer clinical monitoring of patients on maintenance therapy or off treatment completely if they have very bulky or symptomatic disease • There are some patients for whom an intermittent strategy may not be appropriate

34. Limitations • Meta-analysis not based on individual patient data

35. Conclusions • Intermittent strategies of administering first line systemic therapies to patients with unresectable metastatic colorectal cancer: • Do not result in a statistically or clinically significant reduction in overall survival and improve or maintain quality of life compared to continuous administration of therapy. • Offer a reasonable alternative to continuous administration and should be part of an informed discussion of treatment options for this group of patients.

36. Supplemental Slides

37. Background • Effective combinations of cytotoxic and biologic agents have improved outcomes for patients with unresectable metastatic colorectal cancer (mCRC) • As patients live longer it also means a longer exposure to the toxicities of systemic therapies.

38. Methods

39. Methods –Study/Trial Design and Quality • Randomized trials were assessed for key methodological characteristics, using information provided in the trial reports. • The following elements were assessed: • generation of allocation sequence • allocation concealment, • blinding • intention-to-treat analysis • withdrawals, loss to follow-up, • funding source • statistical power calculations, • length of follow-up • differences in baseline patient characteristics, and • early termination.

40. Results

41. Results –Study/Trial Design and Quality

42. Results – Toxicity Summary

43. Results - Quality of Life • Only two of the trials identified for this systematic review reported quality of life (QOL) data : Maughan et al and Adams et al

44. Results - Quality of Life(Maughan) • Maughan et al. administered the Hospital Anxiety and Depression Scale (HADS) and the European Organization for Research Treatment of Cancer (EORTC) Quality of Life Core Questionnaire (QLQ C30) as well as six other questions to trial patients prior to randomization and every six weeks thereafter. • Patients in the intermittent and continuous chemotherapy arms of the trial were similar with respect to physical functioning, overall health, general symptoms and psychological distress

45. Results - Quality of Life(Adams) • Adams et al. administered the EORTC QLQ C30 as well as five other questions to patients at baseline, six weeks, 12 weeks and every 12 weeks thereafter. • QOL questionnaires were completed by less than 67% of patients.

46. Results - Quality of Life(Adams) • At baseline there were no significant differences between patients in the two arms of the trial. • At 12 weeks: • significant detriment with respect to problems in eating or drinking for those in the intermittent chemotherapy arm (OR, 1.23; 95%CI, 1.00-1.50, p=0.045). • there was a significant detriment for intermittent chemotherapy for pain (OR, 1.38; 95%CI, 1.16-1.64), p=0.00029)

47. Results - Quality of Life(Adams) • At 24 weeks, there were significant benefits for the intermittent chemotherapy arm for role functioning (OR, 0.82; 95%CI, 0.70-0.96, p=0.015) and social functioning (OR, 0.82; 95%CI, 0.70-0.96, p=0.016). • detriment for intermittent chemotherapy for pain seen at 12 weeks was not evident at 24 weeks • At this time point there were also significant benefits for the intermittent chemotherapy arm for several symptom scales including: fatigue, nausea and vomiting, appetite loss, constipation, diarrhea, dry or sore mouth, eating or drinking problems, problems handling small objects, and treatment interfering with activities of daily living (all p<0.04).

48. Results – Grade 3/4 Hematologic Toxicity • Similarrates of anemiawith both chemotherapy strategies in all the trials that reported this outcome. • More cases of neutropenia with continuous chemotherapy in Tournigand et al and Chibaudel et al although it was only significant in Tournigand et al • Febrile neutropenia rates were only reported for Tveit et al and rates were similar in both arms • Of the five trials that reported thrombocytopenia: • the incidence was similar for both chemotherapy strategies in three of the trials (Maughan, Labianca and Tveit et al) • nonsignificantlyincreased in the continuous chemotherapy arm of Chibaudel et al • significantly increased for the intermittent chemotherapy arm in Tournigand et al See table in supplementary slides for full results

49. Results – Grade 3/4 Non-Hematologic Toxicity • Incidence of grade 3-4 non-hematologic toxicities tended to be similar for both chemotherapy strategies, with the following exceptions: • Intermittent strategy resulted in significantly more nausea/vomiting and mucositisin Tournigand et al and hand-foot syndrome/rash in Tournigand et al and Tveit et al • The continuous chemotherapy strategy resulted in significantly more cases of the following outcomes in the trials noted: fatigue, neurologic toxicity and hand-foot syndrome in Diaz-Rubio et al See table in supplementary slides for full results