Third- and Further-line Therapy in Advanced NSCLC Patients
Third- and Further-line Therapy in Advanced NSCLC Patients
We have divided targeted therapies into four groups on the basis of their biological mechanisms.
The most widely studied targeted therapies for the third- and further-line treatment of NSCLC are EGFR inhibitors: gefitinib, erlotinib, cetuximab and afatinib (Table 3).
The most important predictive factor for tyrosine kinase inhibitors such as gefitinib, erlotinib and afatinib is represented by the presence of an EGFR-activating mutation. Several randomized trials in both the first- and second-line settings showed that patients with one of these mutations have a greater benefit compared with those who lack this gene signature.
Before these experiences, tyrosine kinase inhibitors had been used after the failure of standard chemotherapy. One of the first articles published on this topic involved 216 NSCLC patients who had been previously treated with at least two lines of chemotherapy who were randomized to receive gefitinib at two dosages (250 or 500 mg/day). Although the authors found no differences in survival between the two drug dosage groups (7 vs 6 months, respectively; p = 0.40), a number of efficacy indices (e.g., symptom improvement, RR, progression and median survival) were strongly correlated with specific clinical patient characteristics (e.g., female gender, never or light smokers and adenocarcinoma features). With regards to toxicity, the group that was treated with the higher dosage had a greater incidence of grade 3/4 side effects, especially diarrhea and skin rash (p = 0.02), and one treatment-related death also occurred due to tumor cavitation with the consequent development of massive hemoptysis.
In a similar experience, Garcia-Velasco et al. published a study on the use of another EGFR inhibitor (erlotinib) in a large group of NSCLC patients who had been previously treated with at least two therapeutic regimens. The authors reported a RR of 11% and SD in 38%, with a median PFS and OS of 3.2 and 5.6 months, respectively. As with Kris and coworkers' experience, these authors found a correlation between high RR and specific clinical patient characteristics. In an attempt to improve efficacy results through the concurrent inhibition of different biological pathways, Dragnev and coworkers evaluated the combination of erlotinib and bexarotene in the third- or further-line treatment of NSCLC. Modest efficacy was observed, probably as result of patient selection. In addition to expected hypertriglyceridemia and skin rash, the authors also observed a higher incidence of other severe side effects (e.g., pulmonary hemorrhage, mouth sores, cough, hypereosinophilic syndrome and abdominal pain), all of which resulted in treatment discontinuation.
Recently, Shao et al. published a retrospective analysis comparing erlotinib and gefitinib in terms of efficacy in third-line therapy. Both EGFR inhibitors yielded the same OS and time of treatment failure, thus confirming the notion that prognostic factors were female gender and histology (adenocarcinoma), independently of the drug used as the second-line treatment.
Conversely, Matsuura et al. did not identify a prognostic role for adenocarcinoma histology in the third-line treatment of advanced NSCLC. In fact, 20 patients without activating EGFR mutations (55% with adenocarcinoma and the remaining 45% with squamous or another NSCLC type) had similar median PFS and OS. A possible explanation for these results might be that the genomic pattern of the patients, all of whom were Asian, made them sensitive to EGFR. It is also possible that the good clinical PS of the patients at the start of the trial influenced this outcome (95% of patients had PS 0–1).
Another specific anti-EGFR therapy is cetuximab, a chimerized antibody of the IgG1 subclass with a fivefold higher affinity for EGFR than natural ligands. Cetuximab stimulates EGFR internalization, effectively removing the receptor from the cell surface for interaction with the ligands. In addition, due to its IgG1 construct, in vitro cetuximab has been shown to mediate antibody-dependent cellular cytotoxicity in certain tumor types, including lung cancer cell lines. On the basis of these preclinical data, Hanna and coworkers published their findings from a Phase II trial in which previously treated NSCLC patients underwent monotherapy with the monoclonal antibody. Notably, approximately 60% of the enrolled patients had previously been treated with at least two systemic regimens. In a series of 66 patients treated with a median of two courses of therapy (range: 1–21), an overall RR of 4.5% was observed, with 30% SD and a DCR of approximately 35%, producing a median PFS of 2.3 months and a median OS of 8.9 months. Notably, all three responding patients did not show EGFR mutations, and hyperexpression of the gene was not related to histologic subtype or smoking history. The most severe side effects were dyspnea, fatigue, infections and anaphylactic reaction. In conclusion, cetuximab is well tolerated in heavily pretreated patients with advanced NSCLC. Efficacy appears to be comparable but not superior to the currently available chemotherapy agents or EGFR tyrosine kinase inhibitors.
Neal and coworkers carried out a clinical study in which cetuximab was tested in 18 heavily pretreated NSCLC patients, the majority of whom had adenocarcinoma subtype and had already received three or more lines of therapy, including EGFR inhibitors. Cetuximab did not induce a tumor response and SD was observed in only five patients (28%), with a median PFS of 1.8 months and a median OS of 7.5 months, irrespective of EGFR or K-RAS mutations. Tolerance was good, with few cases of grade 3/4 toxicities (two interstitial lung diseases, one chest pain and one wheezing). The authors concluded that cetuximab does not yield a response in heavily pretreated (including EGFR inhibitors) patients and suggested that it would be useful to identify predictive factors in order to select patients who could benefit from cetuximab.
The newest EGFR inhibitor is afatinib (BIBW 2992), a novel generation of an irreversible tyrosine kinase inhibitor that selectively targets EGFR and HER1. It has been seen that the irreversible binding of afatinib overcomes the resistance to erlotinib and/or gefitinib, especially when the T790M mutation is present. Murakami and coworkers assessed the safety of afatinib in in a Phase I–II trial of NSCLC patients who had been previously treated with chemotherapy and erlotinib or gefitinib: 13 patients were selected to receive afatinib at a dose ranging from 20 to 50 mg/day. All patients received antidiarrheal drugs in order to prevent gastrointestinal toxicity. The median duration of treatment was 69 days (range: 28–370 days) and the best overall response was SD in nine patients (69%). Grade 3 side effects included diarrhea (one patient) and mucositis (one). The authors concluded by proposing a starting dose of afatinib 50 mg once daily with a tolerability-adapted dose-reduction scheme. This drug dosage was administered in the experimental arm of a Phase IIb/III trial in which afatinib was compared with placebo for the treatment of 585 advanced adenocarcinoma patients who had been previously treated with several lines of therapy, including chemotherapy and reversible EGFR inhibitors. The majority of the enrolled patients had good PS, were of Asian origin and were nonsmokers. The experimental drug produced an RR of approximately 7% and improved PFS over placebo (3.3 vs 1.1 months, respectively; p < 0.0001), but not OS (10.8 vs 12.0 months; p = 0.74). A possible explanation for such results could lie in patient characteristics. Indeed, the majority of enrolled patients – especially those in the placebo arm – had good PS, which allowed them to receive further treatment lines. However, post hoc analysis revealed that patients with mutated EGFR who had previously received erlotinib or gefitinib had better results in terms of PFS and OS. Few cases of grade 3/4 toxicity were reported (grade 3 diarrhea [17%] and grade 3 rash or acne [14%]). Such results could form the basis of a promising new strategy in which EGFR inhibitors are used to rechallenge first-line therapy. In fact, in a retrospective analysis of 14 NSCLC patients harboring EGFR mutations who relapsed after first-line therapy and were retreated with erlotinib, this strategy yielded 36% PR, 50% SD and only 14% PD, with a median PFS of 6.5 months. Notably, five (36%) of these patients had T790M mutations, and these showed two PRs, one SD and two PDs.
Summary. Overall, the effective results in further-line therapy yielded by EGFR inhibitors are modest, but clinical (never smoking and female), histologic (adenocarcinoma features) and biologic (EGFR mutation) parameters may be positive predictive factors in order to select patients who may benefit more from these therapies. The side effects are manageable and consist of rash or acne and diarrhea.
Several inhibitors of angiogenesis have been investigated in advanced NSCLC, as shown in Table 4. Leighl and coworkers conducted a clinical study on the effect of aflibercept (VEGF trap) in 99 heavily pretreated nonsquamous NSCLC patients who had developed resistance to cisplatin and erlotinib. Aflibercept, a recombinant fusion protein, is known to block the activity of VEGF-A, with a consequent inhibition of angiogenesis and tumor cell apoptosis. In a prospective multicenter Phase II trial, this experimental drug produced SD for at least 30 days in 67% of patients, with a median TTP and OS of 2.7 and 6.2 months, respectively. The most important aspect of Leighl et al.'s experience was the safety profile of the drug. As expected, the most common grade 3/4 toxicities were hypertension (23%), dyspnea (21%) and protenuria (10%), but two cases of fatal treatment-related hemoptysis were also registered.
Vandetanib is an oral tyrosine kinase inhibitor that potently inhibits VEGFR-2, VEGFR-3 and EGFR. In 2011, Natale and coworkers carried out a randomized Phase III trial comparing vandetanib with erlotinib in 1240 previously treated NSCLC patients. Notably, approximately a third of the enrolled patients had received two or three prior chemotherapy regimens. The efficacy analysis of the two arms showed similar RRs (12%) and DCRs (41 vs 39%, respectively; p = 0.58). OS analysis confirmed the nonsuperiority of vandetanib over erlotinib (6.8 vs 7.7 months, respectively; p = 0.613). Safety and tolerability profiles showed that patients who received vandetanib had a greater incidence of serious side effects than those who were given erlotinib (50 vs 40%) and a higher rate of discontinuation (14 vs 7%). The most frequent toxicities were rash, pneumonia, dyspnea, diarrhea and hypertension. Lee and coworkers reported a similar experience in which patients who had been previously treated with one or two chemotherapeutic regimens and who had failed EGFR inhibitor therapy were randomized to receive the experimental drug or placebo. Once again, vandetanib failed to improve OS compared with placebo. Patients receiving the experimental drug experienced diarrhea, rash, hypertension and QTc prolongation and required a dose reduction or treatment interruption.
Sorafenib is an oral multikinase inhibitor of angiogenesis and cellular proliferation, with its anticancer effect deriving from the inhibition of VEGFR pathways (VEGFR-2 and -3), PDGFR-β, Raf-1, Flt-3 and c-kit. In a pilot study, sorafenib demonstrated activity in K-RAS-mutated NSCLC, as shown by the fact that three out of the ten enrolled patients achieved PR. In another Phase II trial, 57 patients with K-RAS mutations receiving sorafenib showed PR in seven cases (12%) and SD in 23 cases (40%), with a PFS of 6 months. In 2009, Blumenschein and coworkers tested sorafenib in 52 NSCLC patients, of whom approximately 30% had received two prior treatments. Although no tumor responses were recorded, approximately 59% of cases had SD for 3.4 months; the median PFS was 2.7 month and the OS was 6.7 months. Treatment was generally well tolerated, with the most significant grade 3 side effects being hand–foot skin reaction in five patients and hypertension in two patients. Only three patients were forced to discontinue therapy because of drug-related toxicities (i.e., hand–foot skin reaction in one case, elevated lipase levels in another and myocardial infarction in the last). In conclusion, sorafenib is well tolerated but shows modest activity as a monotherapy in NSCLC.
Sorafenib has been used in combination with EGFR inhibitors such as gefitinib or erlotinib in an attempt to improve efficacy results by inhibiting several biological pathways. In 2007, Adjei and coworkers tested sorafenib in association with gefitinib in a Phase I dose-escalation trial in which 13 (41%) of the enrolled patients had received two or three prior systemic therapies. Treatment was generally well tolerated with few grade 3/4 side effects, specifically dermatologic toxicity (four patients), diarrhea (three patients), elevated hepatic index (five patients) and neurotoxicity (one patient). All of the patients were evaluated for tumor response: one patient with adenocarcinoma had a PR after three prior systemic lines of treatment, whereas 20 patients showed SD; the overall median PFS was 19 weeks. The most important clinical experience in which sorafenib was tested with an EGFR inhibitor (erlotinib) was Spigel and coworkers' trial. In this trial, 166 patients, 65 (42%) of whom had previously received two lines of therapy, were randomly assigned (2:1) to receive erlotinib combined with sorafenib or placebo. The primary end points of the study were RR and PFS, while secondary end points were DCR and OS. The primary end points were not reached and OS was similar in both arms. Such results may be attributable to the small size of the study and some imbalances in patient characteristics at randomization (gender, PS and previous number of lines of treatments). Notably, a post-study analysis showed an increased efficacy of the experimental association with respect to standard treatment in patients with wild-type EGFR, probably because sorafenib may interfere with several biological pathways, thereby blocking tumor growth. Patients treated with the combination erlotinib/sorafenib experienced severe side effects more frequently, particularly diarrhea and hypertension, resulting in more frequent dose reductions or treatment interruptions.
A similar drug is sunitinib malate, an oral selective multitarget tyrosine kinase inhibitor with antiangiogenic and antitumor properties. Sunitinib inhibits VEGFR-1, -2 and -3, PDGFR-α, PDGFR-β and several other related tyrosine kinases and has demonstrated an acceptable profile and clinical efficacy in the treatment of renal cell carcinoma and gastrointestinal stromal tumors. In 2008, Socinski and coworkers tested sunitinib in 60 NSCLC patients after the failure of one or two prior systemic therapeutic regimens, including an EGFR inhibitor. The majority of patients in this Phase II trial had PS 0–1 (60%) and adenocarcinoma subtype (64%). All patients received sunitinib at a dose of 50 mg/day in 6-week cycles comprising once-daily treatment for 4 consecutive weeks followed by 2 weeks of no treatment (schedule 4/2). Seven patients (11%) had a PR and 18 patients (29%) showed SD for at least 8 weeks. The most significant grade 3/4 nonhematological side effects were fatigue/asthenia, pain/myalgia, dyspnea and nausea/vomiting. Notably, there were four cases (7%) of severe or fatal hemorrhage that were suspected to be treatment related. Hence, the most significant hematological side effects were lymphopenia and thrombocytopenia. Transient dose interruption was required in 15 patients (24%), while 14 patients (22%) underwent a dose reduction and 17 patients (29%) permanently discontinued the treatment. The authors concluded that, although sunitinib seems to be effective against heavily pretreated NSCLC, a lower drug dosage would probably improve tolerance.
In order to confirm this hypothesis, Novello and coworkers conducted a study in which patients with similar clinical characteristics to those of Socinski et al.'s experience received sunitinib at a lower dose (37.5 mg/day) in a continuous daily dose schedule. Dose interruption or modification was reported in only 14 patients (30%). The most frequent grade 3/4 treatment-related adverse events were asthenia/fatigue in eight patients (17%) and hypertension in four patients (9%). There was one PR (2%) and 11 SDs (23%), with a median PFS of 2.7 months and an OS of 8.6 months. The authors concluded that sunitinib administered with this modified schedule and dosage was well tolerated and produced interesting efficacy results.
Recently, Scagliotti and coworkers reported the final results of a Phase III randomized trial in which patients who had been previously treated with one or two chemotherapeutic regimens were randomized to receive erlotinib associated with placebo or sunitinib using the modified schedule. Randomization was stratified by smoking history, previous therapy with bevacizumab and EGFR status as determined by immunohistochemistry or FISH (positive vs negative vs unknown). A total of 960 patients were enrolled, of whom about 30% per arm had previously received two or three chemotherapeutic regimens. Severe side effects were more frequent in the combination arm, particularly asthenia/fatigue and diarrhea/vomiting. The experimental arm produced a statistical difference in terms of overall RR (11 vs 7%; p = 0.05) and PFS (p = 0.0023), but not in terms of OS (p = 0.14). Finally, forest plot analysis of prognostic factors did not reveal a specific subgroup that was more likely to benefit from the drug combination.
Summary. VEGFR inhibitors in advanced-setting therapy seem to be not as effective as rescue therapies both as monotherapies and when associated with other types of targeted therapies. Furthermore, these types of therapies have been burdened by side effects, even though they only occurred in a small proportion of the patients who were analyzed.
Bexarotene is a synthetic selective retinoid receptor modulator of numerous intracellular processes, including apoptosis. In Govindan and coworkers' trial, patients who had previously undergone at least two systemic treatments (including a platinum derivate and a taxane) and with PS 0–2 received oral bexarotene in association with levothyroxine and a lipid-lowering agent. Overall, 146 patients were assessable for efficacy and safety (Table 5). The experimental therapy produced modest results in terms of RR (1%), with a median PFS and OS of 2 and 5 months, respectively. The authors attributed these results to the fact that approximately 14% of patients withdrew from the study following a rapid worsening of PS, thus undergoing a limited period of therapy. With regards to the safety profile, the most frequent grade 3/4 side effects were hypertriglyceridemia (23%), dyspnea (16%) and fatigue (10%). Although we cannot be sure whether bexarotene, despite its prophylactic antilipemic therapy, induced the skin reactions that are often associated with modification of the lipemic profile, the dyspnea and fatigue observed in these patients were probably related to lung cancer rather than to bexarotene.
In 2008, Oh and coworkers administered enzastaurin as a second- or third-line therapy in 42% of the 55 NSCLC patients enrolled in their study. The biological antitumor effect of this experimental drug is based on serine/threonine kinase inhibition with induction of tumor cell apoptosis and suppression of angiogenesis. Enzastaurin was administered orally with overall good compliance, except for a few cases of grade 3 toxicity, including fatigue (one patient), ataxia (one patient), pulmonary embolism (one patient) and anemia (one patient). The efficacy results were modest: no objective responses were observed and PFS was 1.8 months, while OS was 8.4 months.
Summary. Multitarget inhibitors in the further-line therapy of NSCLC seem to be ineffective and characterized by severe side effects.
Mapatumumab is a recombinant, high-affinity fully human agonist IgG monoclonal antibody that is specific for human TRAIL-R1, a protein that is overexpressed in a broad range of solid tumors, including lung cancer. By binding the TRAIL-R1 membrane protein, this antibody reduces cell viability, inducing cell death. In 2008, Greco and coworkers tested this drug in 32 heavily pretreated NSCLC patients: 66% had received at least three prior lines of therapy and all had received first-line therapy with a platinum derivate, which was often followed by second-line treatment with taxanes (81%). On average, patients received two cycles of the experimental drug. There were no adverse events requiring the discontinuation of the study agent. The most serious nonhematological side effects were diarrhea, pneumonia, dehydration and pleural effusion, occurring in two patients. Grade 3/4 lymphopenia was observed in seven patients (22%). It is not known exactly how mapatumumab reduces lymphocyte counts, but such changes in these patients were intermittent, thus not increasing the risk of infection. The experimental drug produced SD in nine patients (29%), with a median PFS of 1.2 months.
Substantially different results in terms of efficacy and tolerability were reported by Kwak and coworkers, who explored the antitumor activity of crizotinib, an oral ATP-competitive selective inhibitor of ALK and MET tyrosine kinases.EML4-ALK is an aberrant fusion gene that encodes a cytoplasmatic chimeric protein with constitutive kinase activity, the incidence of which is approximately 2–7% in NSCLC (prevalently adenocarcinomas in young, never or light smokers). After screening approximately 1500 patients for ALK rearrangement with FISH, 82 patients were identified as having this genomic feature. In this study population, crizotinib produced a RR in 47 patients (57%) and SD in 27 patients (33%) for an overall DCR of approximately 90%, with a 6-month PFS of 72%. The median OS for this study has not yet been reached. Notably, 49 patients (59%) had received two or more previous lines of systemic treatment. Overall, the treatment was well tolerated and the most significant grade 3/4 toxicity was hepatic index elevation in approximately 11% of patients.
Recently, Ramalingam et al. tested docetaxel in combination with everolimus, an inhibitor of mTOR, for the second- and third-line therapy of advanced NSCLC. Of the 28 enrolled patients, 12 (43%) had previously received treatment with two lines of chemotherapy. Overall, two PRs (7%) and 15 SDs (54%) were reported, for an overall DCR of 61%, while the median PFS and OS were 4.4 and 9.2 months, respectively. Grade 3/4 side effects were modest but included neutropenia (14%), hyperglycemia (14%) and anemia (11%). The authors concluded that high levels of pAkt, a kinase involved in the mTOR pathways, may be predictive of response to everolimus-containing regimens.
Summary. Other targeted therapies are not effective in advanced-setting NSCLC, with the exception of crizotinib, which is able to produce dramatic responses in NSCLC with EML4-ALK aberrant fusion genes.
Targeted Therapies
We have divided targeted therapies into four groups on the basis of their biological mechanisms.
EGFR Inhibitors
The most widely studied targeted therapies for the third- and further-line treatment of NSCLC are EGFR inhibitors: gefitinib, erlotinib, cetuximab and afatinib (Table 3).
The most important predictive factor for tyrosine kinase inhibitors such as gefitinib, erlotinib and afatinib is represented by the presence of an EGFR-activating mutation. Several randomized trials in both the first- and second-line settings showed that patients with one of these mutations have a greater benefit compared with those who lack this gene signature.
Before these experiences, tyrosine kinase inhibitors had been used after the failure of standard chemotherapy. One of the first articles published on this topic involved 216 NSCLC patients who had been previously treated with at least two lines of chemotherapy who were randomized to receive gefitinib at two dosages (250 or 500 mg/day). Although the authors found no differences in survival between the two drug dosage groups (7 vs 6 months, respectively; p = 0.40), a number of efficacy indices (e.g., symptom improvement, RR, progression and median survival) were strongly correlated with specific clinical patient characteristics (e.g., female gender, never or light smokers and adenocarcinoma features). With regards to toxicity, the group that was treated with the higher dosage had a greater incidence of grade 3/4 side effects, especially diarrhea and skin rash (p = 0.02), and one treatment-related death also occurred due to tumor cavitation with the consequent development of massive hemoptysis.
In a similar experience, Garcia-Velasco et al. published a study on the use of another EGFR inhibitor (erlotinib) in a large group of NSCLC patients who had been previously treated with at least two therapeutic regimens. The authors reported a RR of 11% and SD in 38%, with a median PFS and OS of 3.2 and 5.6 months, respectively. As with Kris and coworkers' experience, these authors found a correlation between high RR and specific clinical patient characteristics. In an attempt to improve efficacy results through the concurrent inhibition of different biological pathways, Dragnev and coworkers evaluated the combination of erlotinib and bexarotene in the third- or further-line treatment of NSCLC. Modest efficacy was observed, probably as result of patient selection. In addition to expected hypertriglyceridemia and skin rash, the authors also observed a higher incidence of other severe side effects (e.g., pulmonary hemorrhage, mouth sores, cough, hypereosinophilic syndrome and abdominal pain), all of which resulted in treatment discontinuation.
Recently, Shao et al. published a retrospective analysis comparing erlotinib and gefitinib in terms of efficacy in third-line therapy. Both EGFR inhibitors yielded the same OS and time of treatment failure, thus confirming the notion that prognostic factors were female gender and histology (adenocarcinoma), independently of the drug used as the second-line treatment.
Conversely, Matsuura et al. did not identify a prognostic role for adenocarcinoma histology in the third-line treatment of advanced NSCLC. In fact, 20 patients without activating EGFR mutations (55% with adenocarcinoma and the remaining 45% with squamous or another NSCLC type) had similar median PFS and OS. A possible explanation for these results might be that the genomic pattern of the patients, all of whom were Asian, made them sensitive to EGFR. It is also possible that the good clinical PS of the patients at the start of the trial influenced this outcome (95% of patients had PS 0–1).
Another specific anti-EGFR therapy is cetuximab, a chimerized antibody of the IgG1 subclass with a fivefold higher affinity for EGFR than natural ligands. Cetuximab stimulates EGFR internalization, effectively removing the receptor from the cell surface for interaction with the ligands. In addition, due to its IgG1 construct, in vitro cetuximab has been shown to mediate antibody-dependent cellular cytotoxicity in certain tumor types, including lung cancer cell lines. On the basis of these preclinical data, Hanna and coworkers published their findings from a Phase II trial in which previously treated NSCLC patients underwent monotherapy with the monoclonal antibody. Notably, approximately 60% of the enrolled patients had previously been treated with at least two systemic regimens. In a series of 66 patients treated with a median of two courses of therapy (range: 1–21), an overall RR of 4.5% was observed, with 30% SD and a DCR of approximately 35%, producing a median PFS of 2.3 months and a median OS of 8.9 months. Notably, all three responding patients did not show EGFR mutations, and hyperexpression of the gene was not related to histologic subtype or smoking history. The most severe side effects were dyspnea, fatigue, infections and anaphylactic reaction. In conclusion, cetuximab is well tolerated in heavily pretreated patients with advanced NSCLC. Efficacy appears to be comparable but not superior to the currently available chemotherapy agents or EGFR tyrosine kinase inhibitors.
Neal and coworkers carried out a clinical study in which cetuximab was tested in 18 heavily pretreated NSCLC patients, the majority of whom had adenocarcinoma subtype and had already received three or more lines of therapy, including EGFR inhibitors. Cetuximab did not induce a tumor response and SD was observed in only five patients (28%), with a median PFS of 1.8 months and a median OS of 7.5 months, irrespective of EGFR or K-RAS mutations. Tolerance was good, with few cases of grade 3/4 toxicities (two interstitial lung diseases, one chest pain and one wheezing). The authors concluded that cetuximab does not yield a response in heavily pretreated (including EGFR inhibitors) patients and suggested that it would be useful to identify predictive factors in order to select patients who could benefit from cetuximab.
The newest EGFR inhibitor is afatinib (BIBW 2992), a novel generation of an irreversible tyrosine kinase inhibitor that selectively targets EGFR and HER1. It has been seen that the irreversible binding of afatinib overcomes the resistance to erlotinib and/or gefitinib, especially when the T790M mutation is present. Murakami and coworkers assessed the safety of afatinib in in a Phase I–II trial of NSCLC patients who had been previously treated with chemotherapy and erlotinib or gefitinib: 13 patients were selected to receive afatinib at a dose ranging from 20 to 50 mg/day. All patients received antidiarrheal drugs in order to prevent gastrointestinal toxicity. The median duration of treatment was 69 days (range: 28–370 days) and the best overall response was SD in nine patients (69%). Grade 3 side effects included diarrhea (one patient) and mucositis (one). The authors concluded by proposing a starting dose of afatinib 50 mg once daily with a tolerability-adapted dose-reduction scheme. This drug dosage was administered in the experimental arm of a Phase IIb/III trial in which afatinib was compared with placebo for the treatment of 585 advanced adenocarcinoma patients who had been previously treated with several lines of therapy, including chemotherapy and reversible EGFR inhibitors. The majority of the enrolled patients had good PS, were of Asian origin and were nonsmokers. The experimental drug produced an RR of approximately 7% and improved PFS over placebo (3.3 vs 1.1 months, respectively; p < 0.0001), but not OS (10.8 vs 12.0 months; p = 0.74). A possible explanation for such results could lie in patient characteristics. Indeed, the majority of enrolled patients – especially those in the placebo arm – had good PS, which allowed them to receive further treatment lines. However, post hoc analysis revealed that patients with mutated EGFR who had previously received erlotinib or gefitinib had better results in terms of PFS and OS. Few cases of grade 3/4 toxicity were reported (grade 3 diarrhea [17%] and grade 3 rash or acne [14%]). Such results could form the basis of a promising new strategy in which EGFR inhibitors are used to rechallenge first-line therapy. In fact, in a retrospective analysis of 14 NSCLC patients harboring EGFR mutations who relapsed after first-line therapy and were retreated with erlotinib, this strategy yielded 36% PR, 50% SD and only 14% PD, with a median PFS of 6.5 months. Notably, five (36%) of these patients had T790M mutations, and these showed two PRs, one SD and two PDs.
Summary. Overall, the effective results in further-line therapy yielded by EGFR inhibitors are modest, but clinical (never smoking and female), histologic (adenocarcinoma features) and biologic (EGFR mutation) parameters may be positive predictive factors in order to select patients who may benefit more from these therapies. The side effects are manageable and consist of rash or acne and diarrhea.
VEGFR Inhibitors
Several inhibitors of angiogenesis have been investigated in advanced NSCLC, as shown in Table 4. Leighl and coworkers conducted a clinical study on the effect of aflibercept (VEGF trap) in 99 heavily pretreated nonsquamous NSCLC patients who had developed resistance to cisplatin and erlotinib. Aflibercept, a recombinant fusion protein, is known to block the activity of VEGF-A, with a consequent inhibition of angiogenesis and tumor cell apoptosis. In a prospective multicenter Phase II trial, this experimental drug produced SD for at least 30 days in 67% of patients, with a median TTP and OS of 2.7 and 6.2 months, respectively. The most important aspect of Leighl et al.'s experience was the safety profile of the drug. As expected, the most common grade 3/4 toxicities were hypertension (23%), dyspnea (21%) and protenuria (10%), but two cases of fatal treatment-related hemoptysis were also registered.
Vandetanib is an oral tyrosine kinase inhibitor that potently inhibits VEGFR-2, VEGFR-3 and EGFR. In 2011, Natale and coworkers carried out a randomized Phase III trial comparing vandetanib with erlotinib in 1240 previously treated NSCLC patients. Notably, approximately a third of the enrolled patients had received two or three prior chemotherapy regimens. The efficacy analysis of the two arms showed similar RRs (12%) and DCRs (41 vs 39%, respectively; p = 0.58). OS analysis confirmed the nonsuperiority of vandetanib over erlotinib (6.8 vs 7.7 months, respectively; p = 0.613). Safety and tolerability profiles showed that patients who received vandetanib had a greater incidence of serious side effects than those who were given erlotinib (50 vs 40%) and a higher rate of discontinuation (14 vs 7%). The most frequent toxicities were rash, pneumonia, dyspnea, diarrhea and hypertension. Lee and coworkers reported a similar experience in which patients who had been previously treated with one or two chemotherapeutic regimens and who had failed EGFR inhibitor therapy were randomized to receive the experimental drug or placebo. Once again, vandetanib failed to improve OS compared with placebo. Patients receiving the experimental drug experienced diarrhea, rash, hypertension and QTc prolongation and required a dose reduction or treatment interruption.
Sorafenib is an oral multikinase inhibitor of angiogenesis and cellular proliferation, with its anticancer effect deriving from the inhibition of VEGFR pathways (VEGFR-2 and -3), PDGFR-β, Raf-1, Flt-3 and c-kit. In a pilot study, sorafenib demonstrated activity in K-RAS-mutated NSCLC, as shown by the fact that three out of the ten enrolled patients achieved PR. In another Phase II trial, 57 patients with K-RAS mutations receiving sorafenib showed PR in seven cases (12%) and SD in 23 cases (40%), with a PFS of 6 months. In 2009, Blumenschein and coworkers tested sorafenib in 52 NSCLC patients, of whom approximately 30% had received two prior treatments. Although no tumor responses were recorded, approximately 59% of cases had SD for 3.4 months; the median PFS was 2.7 month and the OS was 6.7 months. Treatment was generally well tolerated, with the most significant grade 3 side effects being hand–foot skin reaction in five patients and hypertension in two patients. Only three patients were forced to discontinue therapy because of drug-related toxicities (i.e., hand–foot skin reaction in one case, elevated lipase levels in another and myocardial infarction in the last). In conclusion, sorafenib is well tolerated but shows modest activity as a monotherapy in NSCLC.
Sorafenib has been used in combination with EGFR inhibitors such as gefitinib or erlotinib in an attempt to improve efficacy results by inhibiting several biological pathways. In 2007, Adjei and coworkers tested sorafenib in association with gefitinib in a Phase I dose-escalation trial in which 13 (41%) of the enrolled patients had received two or three prior systemic therapies. Treatment was generally well tolerated with few grade 3/4 side effects, specifically dermatologic toxicity (four patients), diarrhea (three patients), elevated hepatic index (five patients) and neurotoxicity (one patient). All of the patients were evaluated for tumor response: one patient with adenocarcinoma had a PR after three prior systemic lines of treatment, whereas 20 patients showed SD; the overall median PFS was 19 weeks. The most important clinical experience in which sorafenib was tested with an EGFR inhibitor (erlotinib) was Spigel and coworkers' trial. In this trial, 166 patients, 65 (42%) of whom had previously received two lines of therapy, were randomly assigned (2:1) to receive erlotinib combined with sorafenib or placebo. The primary end points of the study were RR and PFS, while secondary end points were DCR and OS. The primary end points were not reached and OS was similar in both arms. Such results may be attributable to the small size of the study and some imbalances in patient characteristics at randomization (gender, PS and previous number of lines of treatments). Notably, a post-study analysis showed an increased efficacy of the experimental association with respect to standard treatment in patients with wild-type EGFR, probably because sorafenib may interfere with several biological pathways, thereby blocking tumor growth. Patients treated with the combination erlotinib/sorafenib experienced severe side effects more frequently, particularly diarrhea and hypertension, resulting in more frequent dose reductions or treatment interruptions.
A similar drug is sunitinib malate, an oral selective multitarget tyrosine kinase inhibitor with antiangiogenic and antitumor properties. Sunitinib inhibits VEGFR-1, -2 and -3, PDGFR-α, PDGFR-β and several other related tyrosine kinases and has demonstrated an acceptable profile and clinical efficacy in the treatment of renal cell carcinoma and gastrointestinal stromal tumors. In 2008, Socinski and coworkers tested sunitinib in 60 NSCLC patients after the failure of one or two prior systemic therapeutic regimens, including an EGFR inhibitor. The majority of patients in this Phase II trial had PS 0–1 (60%) and adenocarcinoma subtype (64%). All patients received sunitinib at a dose of 50 mg/day in 6-week cycles comprising once-daily treatment for 4 consecutive weeks followed by 2 weeks of no treatment (schedule 4/2). Seven patients (11%) had a PR and 18 patients (29%) showed SD for at least 8 weeks. The most significant grade 3/4 nonhematological side effects were fatigue/asthenia, pain/myalgia, dyspnea and nausea/vomiting. Notably, there were four cases (7%) of severe or fatal hemorrhage that were suspected to be treatment related. Hence, the most significant hematological side effects were lymphopenia and thrombocytopenia. Transient dose interruption was required in 15 patients (24%), while 14 patients (22%) underwent a dose reduction and 17 patients (29%) permanently discontinued the treatment. The authors concluded that, although sunitinib seems to be effective against heavily pretreated NSCLC, a lower drug dosage would probably improve tolerance.
In order to confirm this hypothesis, Novello and coworkers conducted a study in which patients with similar clinical characteristics to those of Socinski et al.'s experience received sunitinib at a lower dose (37.5 mg/day) in a continuous daily dose schedule. Dose interruption or modification was reported in only 14 patients (30%). The most frequent grade 3/4 treatment-related adverse events were asthenia/fatigue in eight patients (17%) and hypertension in four patients (9%). There was one PR (2%) and 11 SDs (23%), with a median PFS of 2.7 months and an OS of 8.6 months. The authors concluded that sunitinib administered with this modified schedule and dosage was well tolerated and produced interesting efficacy results.
Recently, Scagliotti and coworkers reported the final results of a Phase III randomized trial in which patients who had been previously treated with one or two chemotherapeutic regimens were randomized to receive erlotinib associated with placebo or sunitinib using the modified schedule. Randomization was stratified by smoking history, previous therapy with bevacizumab and EGFR status as determined by immunohistochemistry or FISH (positive vs negative vs unknown). A total of 960 patients were enrolled, of whom about 30% per arm had previously received two or three chemotherapeutic regimens. Severe side effects were more frequent in the combination arm, particularly asthenia/fatigue and diarrhea/vomiting. The experimental arm produced a statistical difference in terms of overall RR (11 vs 7%; p = 0.05) and PFS (p = 0.0023), but not in terms of OS (p = 0.14). Finally, forest plot analysis of prognostic factors did not reveal a specific subgroup that was more likely to benefit from the drug combination.
Summary. VEGFR inhibitors in advanced-setting therapy seem to be not as effective as rescue therapies both as monotherapies and when associated with other types of targeted therapies. Furthermore, these types of therapies have been burdened by side effects, even though they only occurred in a small proportion of the patients who were analyzed.
Multitarget Inhibitors
Bexarotene is a synthetic selective retinoid receptor modulator of numerous intracellular processes, including apoptosis. In Govindan and coworkers' trial, patients who had previously undergone at least two systemic treatments (including a platinum derivate and a taxane) and with PS 0–2 received oral bexarotene in association with levothyroxine and a lipid-lowering agent. Overall, 146 patients were assessable for efficacy and safety (Table 5). The experimental therapy produced modest results in terms of RR (1%), with a median PFS and OS of 2 and 5 months, respectively. The authors attributed these results to the fact that approximately 14% of patients withdrew from the study following a rapid worsening of PS, thus undergoing a limited period of therapy. With regards to the safety profile, the most frequent grade 3/4 side effects were hypertriglyceridemia (23%), dyspnea (16%) and fatigue (10%). Although we cannot be sure whether bexarotene, despite its prophylactic antilipemic therapy, induced the skin reactions that are often associated with modification of the lipemic profile, the dyspnea and fatigue observed in these patients were probably related to lung cancer rather than to bexarotene.
In 2008, Oh and coworkers administered enzastaurin as a second- or third-line therapy in 42% of the 55 NSCLC patients enrolled in their study. The biological antitumor effect of this experimental drug is based on serine/threonine kinase inhibition with induction of tumor cell apoptosis and suppression of angiogenesis. Enzastaurin was administered orally with overall good compliance, except for a few cases of grade 3 toxicity, including fatigue (one patient), ataxia (one patient), pulmonary embolism (one patient) and anemia (one patient). The efficacy results were modest: no objective responses were observed and PFS was 1.8 months, while OS was 8.4 months.
Summary. Multitarget inhibitors in the further-line therapy of NSCLC seem to be ineffective and characterized by severe side effects.
Other Therapies
Mapatumumab is a recombinant, high-affinity fully human agonist IgG monoclonal antibody that is specific for human TRAIL-R1, a protein that is overexpressed in a broad range of solid tumors, including lung cancer. By binding the TRAIL-R1 membrane protein, this antibody reduces cell viability, inducing cell death. In 2008, Greco and coworkers tested this drug in 32 heavily pretreated NSCLC patients: 66% had received at least three prior lines of therapy and all had received first-line therapy with a platinum derivate, which was often followed by second-line treatment with taxanes (81%). On average, patients received two cycles of the experimental drug. There were no adverse events requiring the discontinuation of the study agent. The most serious nonhematological side effects were diarrhea, pneumonia, dehydration and pleural effusion, occurring in two patients. Grade 3/4 lymphopenia was observed in seven patients (22%). It is not known exactly how mapatumumab reduces lymphocyte counts, but such changes in these patients were intermittent, thus not increasing the risk of infection. The experimental drug produced SD in nine patients (29%), with a median PFS of 1.2 months.
Substantially different results in terms of efficacy and tolerability were reported by Kwak and coworkers, who explored the antitumor activity of crizotinib, an oral ATP-competitive selective inhibitor of ALK and MET tyrosine kinases.EML4-ALK is an aberrant fusion gene that encodes a cytoplasmatic chimeric protein with constitutive kinase activity, the incidence of which is approximately 2–7% in NSCLC (prevalently adenocarcinomas in young, never or light smokers). After screening approximately 1500 patients for ALK rearrangement with FISH, 82 patients were identified as having this genomic feature. In this study population, crizotinib produced a RR in 47 patients (57%) and SD in 27 patients (33%) for an overall DCR of approximately 90%, with a 6-month PFS of 72%. The median OS for this study has not yet been reached. Notably, 49 patients (59%) had received two or more previous lines of systemic treatment. Overall, the treatment was well tolerated and the most significant grade 3/4 toxicity was hepatic index elevation in approximately 11% of patients.
Recently, Ramalingam et al. tested docetaxel in combination with everolimus, an inhibitor of mTOR, for the second- and third-line therapy of advanced NSCLC. Of the 28 enrolled patients, 12 (43%) had previously received treatment with two lines of chemotherapy. Overall, two PRs (7%) and 15 SDs (54%) were reported, for an overall DCR of 61%, while the median PFS and OS were 4.4 and 9.2 months, respectively. Grade 3/4 side effects were modest but included neutropenia (14%), hyperglycemia (14%) and anemia (11%). The authors concluded that high levels of pAkt, a kinase involved in the mTOR pathways, may be predictive of response to everolimus-containing regimens.
Summary. Other targeted therapies are not effective in advanced-setting NSCLC, with the exception of crizotinib, which is able to produce dramatic responses in NSCLC with EML4-ALK aberrant fusion genes.
Source...