The OBR Blog

June 02, 2013 - 05:06 pm Posted in ASCO Conference Coverage comments0 Comments

The role of Programmed Death-1 (PD-1) in suppression of antitumor immunity was one of the hottest topics at the 2012 American Society of Clinical Oncology (ASCO) annual meeting and excitement around this topic has only intensified at ASCO 2013 as this story continues to unfold. PD-1 is a key immune checkpoint receptor expressed on activated T-cells and binding of PD-1 to its ligand, PD-L1, results in suppression of the immune response. While the PD-1 pathway normally plays a protective role by attenuating immune-mediated destruction of healthy tissue, the pathway can be exploited by cancer cells to protect themselves from tumor-specific T-cells. PD-L1, frequently expressed on the surface of tumor cells, binds PD-1 and suppresses the anti-tumor immune response. A number of anti-cancer agents that block the interaction between PD-1 and PD-L1 are now in the clinic. The well-attended clinical science symposium on June 2 entitled “PD1/PDL1: An Effective Target in Melanoma,” featured key advances in the clinical development of several agents.

Merck & Company’s lambrolizumab (MK3475), a humanized monoclonal IgG4 antibody against PD-1 recently made headlines when it received the FDA’s newly debuted breakthrough medication status. Results were presented on June 2 from an ongoing phase Ib study designed to explore the safety and clinical activity of lambrolizumab in patients with advanced melanoma (Ribas et al., Abstract 9009). Patients were administered lambrolizumab every two or three weeks until disease progression or unacceptable toxicity. Data from 135 patients were presented. An independent radiologic review indicated a confirmed overall response rate (ORR) of 38% for all patients; however, at the highest dose level (10 mg/kg) the ORR was 52%, including a 10% complete response rate. Response to lambrolizumab was independent of prior Yervoy® (ipilimumab, Bristol-Myers Squibb) treatment. The most common drug-related AEs, (mostly grade 1/2) included fatigue (30.4%), rash (20.7%), and pruritus (20.7%). The incidence of drug-related grade 3/4 AEs was 12.6% and included transaminase elevations, renal insufficiency and hypothyroidism. Drug-related cases of pneumonitis (all grade 1/2) were reported in 6 patients (4.4%), and hypothyroidism was reported in 11 (8.1%) (grade 3, n=1; the remainder grade 1/2).

Roche also showcased its entrant in this race, MPDL3280A, a human monoclonal antibody that targets the ligand PD-L1 and inhibits binding to the PD-1 receptor (Hamid et al., Abstract 9010). Safety and efficacy of MPDL3280A was evaluated in patients with locally advanced or metastatic melanoma. Thirty eight patients were evaluable for efficacy. An ORR of 29% was observed (as measured by RECIST). A potential biomarker for response to this class of agents is overexpression of PD-L1. In this study, PD-L1 expression status was determined by immunohistochemistry (IHC) in 30 patients, and ORR in patients with PD-L1-positive tumors was 27% while ORR in patients with PD-L1 negative tumors was 20%. Stable disease was achieved in 87% of patients with PD-L1 positive tumors compared to 20% of patients with PD-L1 negative tumors. Twenty four-week progression-free survival (PFS) in all melanoma patients was 43%. Several additional patients had delayed antitumor activity after apparent radiographic progression and were counted as progressive disease for the above analyses. Forty four patients were evaluable for safety and the most common drug-related AEs, mostly grade 1/2, included fatigue (59%), headache (32%), and diarrhea (30%). The incidence of drug-related grade 3/4 AEs was 36% and included hyperglycemia (9%), elevated ALT (7%), elevated AST (5%) and back pain (5%); no pneumonitis or colitis were observed.

It was previously demonstrated that Bristol-Myers Squibb’s nivolumab, an IgG4 fully human monoclonal antibody against PD-1, is active in metastatic melanoma, renal cell carcinoma and non-small cell lung cancer. On June 2, data were presented from a study that evaluated the combination of nivolumab plus a multi-peptide vaccine administered to patients with unresectable melanoma who failed at least one regimen for metastatic disease (Webber et al., Abstract 9011). Patients were eligible regardless of whether or not they had received prior Yervoy therapy. Nivolumab was administered (1, 3 or 10 mg/kg) IV every two weeks with subcutaneous vaccine administered before each infusion. The ORR for nivolumab plus peptide vaccine was 25% in Yervoy-refractory patients and 24% in Yervoy-naïve patients, with a clinical benefit rate of 46% and 44%, respectively. Durable responses (up to 33 months) were achieved in both Yervoy-naïve and Yervoy-refractory patients. PD-L1 status was determined by IHC. Using the reported cut-off of ≥5% staining as PD-L1-positive, there was a high response rate of 67% in all patients, although responses were also observed in patients whose tumors were PD-L1-negative (<5% staining: 19%). Common adverse events (all grades) included fatigue, injection site reaction, and pruritus and there were few Grade 3/4 adverse events (rash, n=2)

The final paper in this clinical science symposium presented preliminary results from the first Phase I study to evaluate simultaneous blockade of both the CTLA4 and PD-1/PD-L1 immune checkpoints (Wolchok et al., Abstract 9012) in patients with advanced melanoma who had received ≤3 prior therapies. A total of 53 patients who were Yervoy-naïve received IV nivolumab and Yervoy concurrently, every 3 weeks for 4 doses, followed by nivolumab alone every 3 weeks for 4 additional doses. At week 24, combined treatment was continued every 12 weeks for 8 more doses in patients with disease control and no dose-limiting toxicities. In two sequenced-regimen cohorts, 33 patients who progressed on prior standard Yervoy therapy were treated with nivolumab (every two weeks for up to 48 doses). The confirmed ORR for patients who received concurrent nivolumab and Yervoy was 40%, the majority of which occurred rapidly and exceeded 80% tumor shrinkage. A 53% confirmed ORR was achieved in patients who received 1 mg/kg nivolumab plus 3 mg/kg Yervoy including 3 complete responses and 6 partial responses. All nine of these patients had ≥80% reduction in tumor volume by 12 weeks or at their first assessment. Grade 3/4 toxicities occurred in 53% of patients who received concurrent nivolumab plus Yervoy. The most common Grade 3/4 toxicities were asymptomatic lab abnormalities including elevated lipase, and elevated AST and ALT. The confirmed ORR was 20% for patients who received nivolumab following progression on Yervoy and Grade 3/4 AEs occurred in 18% of patients who received nivolumab subsequent to Yervoy; asymptomatic elevated lipase was the most common Grade 3/4 AE (n=2, 6%). The 1 mg/kg nivolumab plus 3 mg/kg Yervoy dose was chosen to move forward into Phase III trials.

Data presented at the 2012 ASCO annual meeting established the potential of targeting the PD-1/PD-L1 checkpoint as a therapeutic strategy. These agents were shown to be very well tolerated with efficacy in several advanced tumor types. However, many questions were left unanswered about this novel approach to targeting cancer. Some of these questions were answered on June 2. Two of the studies presented demonstrated that progression of melanoma while on Yervoy therapy does not impact efficacy of subsequent PD-1/PD-L1 blockade. In 2012, data was presented that suggested that expression of PD-L1 by tumor cells may serve as a biomarker to identify patients who are likely to respond to anti-PD-1 therapy and those who will not respond. Today we learned that although PD-L1 expression correlates with response, it does not predict response. While PD-L1-positive tumors have a higher rate of disease control, responses are also achieved in patients with PD-L1-negative tumors, albeit at lower rates. Finally, data presented on June 2 demonstrated the feasibility of concurrent blockade of multiple immune checkpoint pathways. Administration of the CTLA4 pathway inhibitor Yervoy plus the PD-1 monoclonal nivolumab resulted in rapid response at rates as high as 53%. The rate, rapidity and depth of responses reported exceed those achieved with either of these agents administered as monotherapy, and the toxicity associated with this combination was manageable. The melanoma market has only recently been transformed, and these data and ongoing clinical development with these agents suggest that it’s set to be transformed even further!

By Stephanie Hawthorne, Director, Kantar Health and Greg Wolfe, Senior Consultant, Kantar Health

June 02, 2013 - 05:06 pm Posted in ASCO Conference Coverage comments1 Comments

Activation of the PI3K-AKT-mTOR pathway has been implicated as one of several potential mechanisms for acquired resistance to HER2-targeted and hormone therapy in breast cancer. Preclinical and clinical results support investigation into whether acquired resistance can be overcome by downstream inhibition of this pathway. A number of agents currently in clinical development target different kinases in the pathway, the most advanced of which are the mTOR inhibitors ― Afinitor® (everolimus, Novartis) and Torisel® (temsirolimus, Pfizer), both approved for various cancer indications. Within breast cancer, Afinitor in combination with Aromasin® (exemestane, Pfizer) was approved by the U.S. FDA and the European Medicines Agency (EMA) in July 2012 for the treatment of HR+ metastatic breast cancer (mBC), in patients who have failed a prior aromatase inhibitor. This approval was based on a 6.9 month improvement in progression-free survival (PFS) and a trend to overall survival (OS) benefit at the interim analysis (HR 0.77)1. Having clearly demonstrated the role of mTOR inhibition in overcoming resistance to hormone therapy, the natural question is whether this approach is clinically effective in patients resistant to HER2-targeted therapy.

To address this question, Novartis initiated the Phase III BOLERO-3 trial (NCT01007942) which compared Herceptin® (trastuzumab, Genentech/Roche) plus vinorelbine with or without Afinitor in 569 locally advanced or metastatic breast cancer patients who had received a prior taxane and who were resistant to Herceptin. Notably, resistance was defined as either progression while on or within 12 months of adjuvant Herceptin or progression within 4 weeks of Herceptin for metastatic disease. The first results of this trial were reported on June 2 at the 2013 American Society of Clinical Oncology (ASCO) annual meeting in the oral session for HER2/ER Breast Cancer2. Ultimately, 16% of patients were enrolled in the first-line setting, 42% were enrolled as second-line, and 27% had received prior Tykerb® (lapatinib, GlaxoSmithKline). The addition of Afinitor significantly improved PFS (the primary endpoint) by local assessment by 1.2 months (7.0 months versus 5.8 months, HR 0.78, p=0.0067). There was no significant difference in any of the reported secondary endpoints for the Afinitor and placebo arms: objective response rate (ORR: 40.8% versus 37.2%, p=0.2108), quality of life (time to definitive deterioration of global health status QL2 score: 8.3 months versus 7.3 months, HR 0.98, p=0.8386), and overall survival (at interim analysis based on 220 of 384 events: 36.3% versus 41.1% deaths; statistics not provided), respectively. Subgroup analyses suggest that the Afinitor benefit was similar regardless of performance status or prior Tykerb exposure, but differential benefits were observed in a few patient subgroups ― patients who received prior neo/adjuvant Herceptin derived greater benefit with Afinitor than those who had not (HR 0.65 versus 0.92), as did patients who were ER-/PR- compared to those who were ER+ and/or PR+ (HR 0.65 versus 0.93), and those who were younger (<65 years HR 0.77 versus ≥65 years HR 0.93).

In terms of safety, the toxicity profile was similar to that reported with Afinitor in other studies, namely higher rates of Grade 3/4 stomatitis, fatigue, hyperglycemia, anemia, and febrile neutropenia. The adverse event profile and the need for dose reductions lead to a decreased dose intensity for Afinitor compared to placebo (77% versus 96% of planned dose intensity; vinorelbine dose intensity was reduced in both arms and not statistically different (64% versus 73%, respectively).

These results are somewhat disappointing. The magnitude of PFS benefit with Afinitor was modest in the intent to treat population (1.2 month difference at the median and a 22% reduction in the risk of progression or death). While statistically significant, the clinical meaningfulness of this benefit is questionable. With regard to the PFS analysis, only investigator-assessed outcomes were reported. An independent central review analysis would be ideal to have since typically the level of benefit varies between investigator and central review analysis; such an analysis is likely underway. The OS analysis is premature, and a Kaplan-Meier curve or any statistical analyses were not presented. Although there is a numerical 4.8% reduction in the number of deaths in favor of the Afinitor arm, the available data is too limited to allow us to make any prediction as to whether the OS data will eventually prove to be significant or not. The final OS data is expected in 2014.

There is a precedent for approval in relapsed breast cancer with a PFS benefit without OS benefit ― Tykerb was approved based on a 4.3 month PFS benefit but no difference in OS3 ― although more recently Kadycla® (ado-trastuzumab, Genentech/Roche) was approved based on a significant improvement in both PFS and OS compared to Tykerb/capecitabine4. Whether or not the data presented on June 2 for Afinitor would support regulatory approval is therefore questionable.

The treatment of HER2+ breast cancer has exploded in the last year, expanding from having just two options ― Herceptin and Tykerb ― to now having four, following the approval of Perjeta® (pertuzumab, Genentech/Roche) by the FDA in June 2012 and the EMA in March 2013, and the approval of Kadcyla by the FDA in February 2013 (European approval still pending). Physicians are just starting to develop their paradigms to incorporate all of these agents into treatment, although the National Comprehensive Cancer Network (NCCN) recommends first-line treatment with Herceptin plus Perjeta and docetaxel followed by Kadcyla in second-line for patients with HER2+ mBC, with other options (such as Tykerb plus capecitabine or Herceptin re-treatment) for later lines5. If approved, where will Afinitor fit into this treatment paradigm? If we compare the outcomes from BOLERO-3 with those from other trials in the Herceptin-pretreated setting (see Table 2), the data suggests that Afinitor may provide a weaker level of benefit than Kadcyla or Tykerb/capecitabine. It’s likely that any role for Afinitor will be limited to third-line or even later, especially since Afinitor proved effective in patients treated with prior Tykerb. Additionally, the suggestion that all clinical benefit with Afinitor was limited to those patients who were ER-/PR- could further limit its utilization to the less than half of HER2+ patients who are ER-/PR-.

Other questions remain. Which dose of Afinitor should be used in breast cancer? The BOLERO-3 trial utilized a 5 mg/day dose, whereas in the BOLERO-2 trial in combination with Aromasin and in the ongoing BOLERO-1 trial in first-line HER2+ mBC Afinitor is given at a dose of 10 mg/day. With the reduced dose intensity observed in BOLERO-3 it seems unlikely that a higher dose should be used, but are there implications for the ongoing BOLERO-1 trial? Should Afinitor be restricted to the subpopulations that derived the greatest benefit (or conversely, not be used in those subgroups who derived no benefit)? Do the data from BOLERO-3 foreshadow outcomes in the ongoing BOLERO-1 trial (NCT00876395), which will compare Herceptin/paclitaxel with or without Afinitor in first-line HER2+ mBC; results are expected later this year.


  1. Afinitor prescribing information, accessed June 2, 2013.
  2. O’Regan et al, Abstract 505, ASCO 2013
  3. Geyer et al, NEJM, 355(26): 2733-2743, 2006.
  4. Kadcyla prescribing information, accessed June 2, 2013.
  5. National Comprehensive Cancer Network guidelines, Breast Cancer, version 3.2013.
  6. Tykerb prescribing information, accessed June 2, 2013.

By Stephanie Hawthorne, Director, Kantar Health and Neesha Suvarna, Consultant, Kantar Health

June 01, 2013 - 07:06 pm Posted in ASCO Conference Coverage comments0 Comments

Although first-line combination chemotherapy with a platinum (carboplatin or cisplatin) plus a taxane (primarily paclitaxel) is very successful at producing high response rates in first-line advanced ovarian cancer patients, the vast majority (up to 70%1) of these patients still experience a recurrence. Based on this high unmet need to extend progression free survival (PFS) in these first-line patients, maintenance therapy in now utilized in some advanced ovarian cancer patients. Maintenance therapy was initially introduced to ovarian cancer in the Phase III SWOG-9701/GOG-178 trial that found a significant PFS benefit of 12 cycles versus 3 cycles of paclitaxel in patients who had received a platinum-paclitaxel regimen as first-line therapy2,3. However, other trials evaluating maintenance chemotherapy in ovarian cancer have shown no benefit4, creating a controversy regarding whether maintenance chemotherapy is effective or not. Several years later, two Phase III studies evaluated the ability of anti-angiogenic agents to improve outcomes when used in the induction and maintenance settings. The GOG-0218 trial5 and the ICON7 trial6, evaluated the addition of Avastin® (bevacizumab, Genentech/Roche) to standard first-line carboplatin-paclitaxel combination chemotherapy along with continued Avastin maintenance therapy compared to carboplatin-paclitaxel chemotherapy alone with no maintenance therapy. Both trials found a significant PFS benefit with Avastin as first-line induction therapy and maintenance therapy, although interim analyses to-date have failed to show an overall survival (OS) benefit.

Although the concept of maintenance therapy in ovarian cancer has now been present for several years, it has not yet been adopted uniformly into practice. According to data from Kantar Health’s CancerMPact® Treatment Architecture, 20137, approximately 75% of patients in the United States and Western Europe who receive Avastin-based induction regimens as first-line therapy (which is approximately 20% of patients) also receive maintenance therapy; the high use of maintenance Avastin in these patients is supported by data from the GOG-0218 trial that showed no PFS or OS benefit if Avastin was given as part of induction only5. In contrast, in patients who receive a non-Avastin-containing induction regimen as first-line therapy (approximately 80% of patients), maintenance therapy is less common, with only 45-55% of those patients currently receiving any maintenance therapy. This leaves a great deal of room in the market for new, efficacious agents in the maintenance setting.

In addition to Avastin, a number of other anti-angiogenic agents have entered development for ovarian cancer, and Votrient ® (pazopanib, GlaxoSmithKline) has entered the ring specifically in the maintenance setting. Votrient is a broad spectrum tyrosine kinase inhibitor that is currently approved in the U.S. and the European Union for use in renal cell carcinoma and soft tissue sarcoma. Like Avastin, it has activity against VEGF receptors, and it also has activity against PDGF receptors and c-kit. In 2009, the Phase III international Intergroup AGO-OVAR16 trial (NCT00866697, also known as the POIZE trial) was initiated to evaluate Votrient as maintenance therapy in patients with advanced ovarian cancer. A total of 940 patients who had not progressed after at least 5 cycles of chemotherapy with a platinum-taxane combination were randomized 1:1 to receive either placebo or maintenance therapy with single agent Votrient (800 mg QD) for up to two years. The primary endpoint of the trial was PFS, and secondary endpoints included OS, 3-year PFS, quality of life measurements, and safety/tolerability.

Results of the trial were presented on June 1at the 2013 American Society of Clinical Oncology (ASCO) annual meeting1, and the trial met its primary endpoint, with Votrient maintenance significantly extending median PFS by 5.6 months compared to placebo (mPFS: 17.9 months versus 12.3 months, HR: 0.766, p<0.0021, see Table 1). However, although mature OS results are not yet available, an interim analysis (with events from only 20.1% of patients available) did not find any survival benefit of the Votrient arm compared to the placebo arm (medians not yet reached in either arm, HR: 0.994). With the current immature OS data, the Kaplan-Meier survival curves appear to be completely overlapping.

Adverse events observed in the Votrient arm were in line with the known safety profile of Votrient from other tumors; the most frequently reported Grade 3/4 adverse events were hypertension, neutropenia, liver-related toxicity and diarrhea (see Table 1). As would be expected, there was a higher incidence of adverse events in the Votrient arm compared to placebo, and patients in the Votrient arm had a higher rate of dose reduction (58% versus 14%). In addition, patients had a shorter mean exposure to Votrient compared to placebo (8.9 months versus 11.7 months).

If the mature OS data from the AVO-OVAR16 trial continue to be non-significant, Votrient is likely to experience the same regulatory fate in ovarian cancer as Avastin. As mentioned earlier, the Phase III trials evaluating Avastin as first-line induction and maintenance therapy found a significant PFS benefit, but not an OS benefit. Based on these results, Avastin was approved by the European Medicines Agency (EMA) in late 2011 for use in first-line induction and maintenance therapy for advanced ovarian cancer. However, without an OS benefit, Roche has not filed for approval in the U.S., strongly suggesting that the FDA has provided guidance that PFS without OS is not approvable. Similarly, if mature OS data for the AVO-OVAR16 trial continue to show no OS benefit for Votrient, it is unlikely to be approved as maintenance therapy by the FDA, even with the significant PFS benefit. However, as with Avastin, it is likely that this PFS benefit alone will support approval of Votrient by the EMA. Even if it does not receive an FDA approval, however, Votrient (again, like Avastin) is approved in other indications in the United States, and it could still see off-label U.S. utilization, especially if it receives an NCCN recommendation. Notably, however, the NCCN has given Avastin a Category 3 recommendation for use in first-line, which suggests substantial disagreement among panel members as to the clinical benefit provided by Avastin. With similar level of benefit for Votrient (discussed below), Votrient is likely to receive a similar low-level recommendation.

Aside from the question of regulatory approval, there is also the question of potential adoption of Votrient in the maintenance setting. Due to different Phase III trial designs, any use of maintenance Votrient is most likely to be restricted to a separate patient population from maintenance Avastin, which is largely limited to patients who receive Avastin-based induction regimens7. In contrast, patients in the AVO-OVAR16 trial who received maintenance Votrient received non-Avastin-containing induction regimens (a platinum-taxane combination), and Votrient is most likely to be used in (and any regulatory approval would be limited to) this patient population. With the majority of patients in the U.S. and Europe currently receiving non-Avastin first-line regimens, this leaves a great deal of room in the market for Votrient to establish itself as a standard choice for maintenance therapy in this patient population.

As can be seen in Table 2, cross trial comparisons suggest that the PFS benefit of maintenance Votrient is roughly comparable to both Avastin induction/maintenance as well as 12-months paclitaxel maintenance; although the magnitude of PFS benefit at the median is greater with Votrient compared to that achieved with Avastin in its two studies (5.6 months compared with 2.4-2.9 months), the Hazard Ratio is similar in across these three studies. Without stand-out efficacy results, Votrient will need to differentiate itself in other ways. As anti-angiogenic agents, both Avastin and Votrient have overlapping toxicity profiles, but there are some distinctions; in ovarian cancer trials, Votrient has shown a higher incidence of hypertension, while Avastin has shown a higher rate of gastrointestinal perforations (see Table 3). Both Avastin and Votrient also have a different toxicity profile from paclitaxel, which can cause sensory neuropathy in patients. Other factors are also likely to affect adoption rates. Quality of life data has yet to be reported for the AVO-OVAR16 trial, but could be a significant factor in willingness to prescribe. Convenience of administration may also play a role ― Votrient is an oral agent while Avastin and standard chemotherapy are administered intravenously, and the prolonged administration of maintenance therapy could favor an oral. Cost will also be a significant factor ― one year of low-dose Avastin maintenance costs approximately $40,000 whereas two years of Votrient maintenance will cost approximately $171,000. This high price tag is sure to hinder use in both the U.S. and the more cost-sensitive European market.


  1. duBois et al., ASCO 2013, Abstract LBA5503.
  2. Markman et al., JCO, 21(13): 2460-2465, 2003.
  3. Markman, ASCO 2006, Abstract 5005.
  4. Pecorelli et al., JCO 27(28): 4642-4648, 2009.
  5. Burger et al., NEJM, 2011.
  6. Perren et al., NEJM, 2011.
  7. Kantar Health, CancerMPact® Treatment Architecture U.S. and Western Europe, accessed June 1, 2013.

By Stephanie Hawthorne, PhD, Director, Kantar Health and Cory Blaiss, Analyst, Kantar Health

June 01, 2013 - 07:06 pm Posted in ASCO Conference Coverage comments0 Comments

Another year and the 2013 American Society of Clinical Oncology (ASCO) conference provides more intriguing clinical data to shape the future of oncology therapy. One of the hottest presentations this year (pun intended) is the results from a German cooperative group-sponsored FIRE-3 study (n=592), comparing first-line FOLFIRI in combination with low-dose (5 mg/kg every two weeks) Avastin (bevacizumab, Roche/Genentech) or standard dose Erbitux (cetuximab; Bristol Myers Squibb, Eli Lilly, and Merck KGaA) in metastatic colorectal cancer (mCRC) patients with KRAS wild-type tumors. Both targeted agents have previously shown to significantly improve PFS compared with chemotherapy alone, and both have demonstrated a strong trend to OS benefit in previous first-line studies. In the absence of comparative data between the two agents, the current practice patterns have evolved based on safety profile, cost of therapy, order of market entry, and KRAS status. With Avastin the primary biologic option in mCRC patients with a KRAS mutation, the question remained which option was superior in KRAS wild-type tumors. Prospective head-to-head studies were required to compare these two agents in order to definitively answer the question of which agent is the better option in first-line mCRC.

The FIRE-3 results did not disappoint but questions remain after some potentially conflicting results (Stinitzing, Abstract LBA3506). The primary endpoint was overall response rate, which was not significantly different in either arm (62.0% with Erbitux and 58% with Avastin, HR=1.18, p=0.183), but there was a trend to benefit in the Erbitux arm, as well as a higher percentage of complete responses (4.4% versus 1.4%). However, the most interesting conundrum was the fact that there was no difference in progression-free survival with the curves being superimposable (PFS: 10.0 months versus 10.3 months, HR=1.06, p=0.547) but there was a statistically significant benefit in median overall survival (mOS: 28.7 months with Erbitux and 25.0 months with Avastin, HR=0.77, p=0.017). It is the OS Kaplan-Meier curve that tells the full story. The curves begin to separate around 18 months, well after the 10-month PFS median, and the curves continue to separate past the median OS where the superiority of Erbitux arm becomes far more pronounced and is evidently durable.

The immediate question becomes: how can an agent that appears similar in terms of response and identical in terms of PFS, generate such a strong OS benefit? The trial was well balanced, there were no new safety signals.  Duration of therapy and rates of liver resection were comparable. Unfortunately, the answer is not yet available though the investigator did offer the possibility of depth of response as an explanation. It is possible that tumor shrinkage was greater with Erbitux, which would not necessarily effect the measurement of PFS but could allow for more effective subsequent lines of therapy due to the smaller tumor size. As evidence, the speaker offered an analysis of the CRYSTAL study that originally evaluated Erbitux in combination with front-line chemotherapy. The analysis was also presented at ASCO this year and suggested tumor volume shrinkage (rather than the strict RECIST response criteria) could predict a strong OS benefit, but not necessarily effect the time to progression (Mansmann, Abstract 3630). A similar analysis will be performed on the FIRE-3 results in the near future.

One final question that remains is: How will the FIRE-3 results influence U.S. and EU physicians? FIRE-3 could be viewed favorably in Europe where first-line use of Erbitux for KRAS wild-type mCRC patients is utilized to a greater extent. FIRE-3 will likely further increase this practice. Conversely in the U.S., Avastin has been highly entrenched as the preferred first-line biologic to use in combination with first-line chemotherapy for nearly a decade. Additionally, U.S. physicians prefer to use FOLFOX in the first line in combination with Avastin, so they may not see the FIRE-3 study as accurately depicting their first-line treatment of choice. However, in late 2013, results are anticipated from the CALGB 80405 study, which has a very similar design as the FIRE-3 study, evaluating Avastin versus Erbitux in the front line colorectal cancer. In this study patients will be allowed to be given FOLFOX or FOLFIRI as the combination chemotherapy, which likely means that this study will be more influential on the U.S. market, compared with FIRE-3. With the CALGB 80405 trial being initiated around the same time as the FIRE-3 study, the oncology community will not have to wait long to digest the full results and implications from both of these studies.

By Gordon Gochenauer, Director, Kantar Health and Josh Garcia  Associate Consultant, Kantar Health

article register

Recent Posts

Recent Comments