Neil Osterweil

BOSTON – Treatment of pneumonitis associated with the PD-1 axis checkpoint inhibitors requires close monitoring of patients and rapid clinical response, an oncologist advised.

“We need to be vigilant when we treat our patients. If a patient has cough or shortness of breath, you take it seriously, even if [it is] a patient who has lung cancer and is a smoker,” said Dr. Scott Gettinger of Yale Cancer Center, New Haven, Conn.

He recommended that before starting patients on a programmed death-1 (PD-1) axis inhibitor such as pembrolizumab (Keytruda) or nivolumab (Opdivo), clinicians get baseline oxygen saturation levels to obtain an objective measure for following patients during therapy.

“When you suspect pneumonitis you have to start steroids right away, or patients can spiral down,” he said at the AACR/NCI/EORTC International Conference on Molecular Targets and Cancer Therapeutics.

Pneumonitis – characterized by cough, dyspnea, and hypoxia – is a common adverse event associated with PD-1 and PD-ligand 1 (PD-L1) inhibitors, but is rarely seen in patients treated with CTLA-4 inhibitors such as ipilimumab (Yervoy). Current evidence suggests that pneumonitis is more prevalent in patients with non–small-cell lung cancer, occurring in approximately 4%-6% of patients cases, than in melanoma (1%). The difference is probably due to a history of smoking common to the majority of patients with NSCLC, Dr. Gettinger said.

“The other theme that we’re beginning to see is that maybe pneumonitis is a bit more common with anti-PD-1 vs. anti-PD-L1 antibodies,” he said.

It’s theorized that PD-1 inhibitors may block binding of PD-L2 to one of its binding partners, thereby interfering with respiratory tolerance, he said.

Evidence from the pivotal clinical trial of nivolumab in lung cancer (Checkmate 057) suggests that the time to onset of pneumonitis was a median of 31.1 weeks (range 11.7 to 56.9 weeks). The pneumonitis resolved in about 5-7 weeks.

Investigators at Memorial Sloan Kettering Cancer Center in New York City reported at the European Cancer Congress 2015 on pneumonitis in 36 of 653 patients treated with an anti PD-1/PD-L1 monoclonal antibody from 2009 through 2014, 33 of whom had received a PD-1 inhibitor, and 3 of whom received a PD-L1 inhibitor.

They found that pneumonitis in patients with lung cancer tended to resemble chronic obstructive pneumonia, whereas patients with melanoma were more likely to have ground-glass opacifications on radiography. There was a trend, falling just short of significance, toward association of COP-like pneumonitis with development of grade 3 or greater, and with a requirement for more than one type of immunosuppression, compared to other subtypes.

Algorithms for management

Dr. Gettinger briefly outlined an algorithm offered by Bristol-Myers Squibb for management of suspected pulmonary toxicity with nivolumab. For management of patients with grade 1 toxicities (asymptomatic, with radiographic changes only), it recommends that clinicians consider delay of immuno-oncologic (I-O) therapy, monitor for symptoms every 2-3 days, and consider consultations with pulmonary and infectious disease specialists.

For grade 2 pneumonitis, marked by mild to moderate new symptoms, the algorithm calls for clinicians to delay I-O therapy per protocol, consult with pulmonary and ID specialists, monitor symptoms daily and consider hospitalization, start the patient on steroids with 1.0 mg/kg per day methylprednisolone IV or the oral equivalent, and consider bronchoscopy and lung biopsy.

For patients with grade 3 or 4 toxicities, marked by severe new symptoms, new or worsening hypoxia, or other life-threatening symptoms, the algorithm states that clinicians should discontinue I-O therapy, hospitalize the patient, consult with pulmonary and ID, give 2-4 mg/kg per day methylprednisolone IV or the oral equivalent, add prophylactic antibiotics for opportunistic infections, and consider bronchoscopy and lung biopsy.

At Yale, Dr. Gettinger and colleagues, when presented with a symptomatic patient on a PD-axis inhibitor, will first rule out other etiologies such as infection, chronic obstructive pulmonary disease exacerbation, or cancer progression), typically with bronchoscopy.

For patients with moderate pneumonitis, they may treat with prednisone for 1 or 2 weeks, with a 4-6 week taper begun as symptoms start to resolve.

“In a patient who has profound hypoxia and shortness of breath, we may want to go higher: We might give them 2 mg/kg twice a day of [methylprednisolone] in the hospital, wait until they get better, and then slowly taper them, whether it be over 4 or 6 weeks or longer. Occasionally our patients need to get even higher doses of steroids and there’s really nothing to guide us. Who knows if 1 gram of [methylprednisolone] may be better than 60 g of prednisone? But we do it, and patients do get better with the higher doses,” he said.

In rare instances, patients may required other immunosuppressive agents, such as infliximab (Remicade), cyclophosphamide, or mycophenolate mofetil.

Patients who require additional immunosuppressive agents to resolve severe pneumonitis tend to have poor outcomes, Dr. Gettinger said.

Re-challenge of patients with a PD-1 or PD-L1 inhibitor following resolution of pneumonitis appears to be inadvisable for all patients except possibly those with grade 1 (asymptomatic) toxicity, due to the high risk of recurrence, he said.

BOSTON – Treatment of pneumonitis associated with the PD-1 axis checkpoint inhibitors requires close monitoring of patients and rapid clinical response, an oncologist advised.

“We need to be vigilant when we treat our patients. If a patient has cough or shortness of breath, you take it seriously, even if [it is] a patient who has lung cancer and is a smoker,” said Dr. Scott Gettinger of Yale Cancer Center, New Haven, Conn.

He recommended that before starting patients on a programmed death-1 (PD-1) axis inhibitor such as pembrolizumab (Keytruda) or nivolumab (Opdivo), clinicians get baseline oxygen saturation levels to obtain an objective measure for following patients during therapy.

“When you suspect pneumonitis you have to start steroids right away, or patients can spiral down,” he said at the AACR/NCI/EORTC International Conference on Molecular Targets and Cancer Therapeutics.

Pneumonitis – characterized by cough, dyspnea, and hypoxia – is a common adverse event associated with PD-1 and PD-ligand 1 (PD-L1) inhibitors, but is rarely seen in patients treated with CTLA-4 inhibitors such as ipilimumab (Yervoy). Current evidence suggests that pneumonitis is more prevalent in patients with non–small-cell lung cancer, occurring in approximately 4%-6% of patients cases, than in melanoma (1%). The difference is probably due to a history of smoking common to the majority of patients with NSCLC, Dr. Gettinger said.

“The other theme that we’re beginning to see is that maybe pneumonitis is a bit more common with anti-PD-1 vs. anti-PD-L1 antibodies,” he said.

It’s theorized that PD-1 inhibitors may block binding of PD-L2 to one of its binding partners, thereby interfering with respiratory tolerance, he said.

Evidence from the pivotal clinical trial of nivolumab in lung cancer (Checkmate 057) suggests that the time to onset of pneumonitis was a median of 31.1 weeks (range 11.7 to 56.9 weeks). The pneumonitis resolved in about 5-7 weeks.

Investigators at Memorial Sloan Kettering Cancer Center in New York City reported at the European Cancer Congress 2015 on pneumonitis in 36 of 653 patients treated with an anti PD-1/PD-L1 monoclonal antibody from 2009 through 2014, 33 of whom had received a PD-1 inhibitor, and 3 of whom received a PD-L1 inhibitor.

They found that pneumonitis in patients with lung cancer tended to resemble chronic obstructive pneumonia, whereas patients with melanoma were more likely to have ground-glass opacifications on radiography. There was a trend, falling just short of significance, toward association of COP-like pneumonitis with development of grade 3 or greater, and with a requirement for more than one type of immunosuppression, compared to other subtypes.

Algorithms for management

Dr. Gettinger briefly outlined an algorithm offered by Bristol-Myers Squibb for management of suspected pulmonary toxicity with nivolumab. For management of patients with grade 1 toxicities (asymptomatic, with radiographic changes only), it recommends that clinicians consider delay of immuno-oncologic (I-O) therapy, monitor for symptoms every 2-3 days, and consider consultations with pulmonary and infectious disease specialists.

For grade 2 pneumonitis, marked by mild to moderate new symptoms, the algorithm calls for clinicians to delay I-O therapy per protocol, consult with pulmonary and ID specialists, monitor symptoms daily and consider hospitalization, start the patient on steroids with 1.0 mg/kg per day methylprednisolone IV or the oral equivalent, and consider bronchoscopy and lung biopsy.

For patients with grade 3 or 4 toxicities, marked by severe new symptoms, new or worsening hypoxia, or other life-threatening symptoms, the algorithm states that clinicians should discontinue I-O therapy, hospitalize the patient, consult with pulmonary and ID, give 2-4 mg/kg per day methylprednisolone IV or the oral equivalent, add prophylactic antibiotics for opportunistic infections, and consider bronchoscopy and lung biopsy.

At Yale, Dr. Gettinger and colleagues, when presented with a symptomatic patient on a PD-axis inhibitor, will first rule out other etiologies such as infection, chronic obstructive pulmonary disease exacerbation, or cancer progression), typically with bronchoscopy.

For patients with moderate pneumonitis, they may treat with prednisone for 1 or 2 weeks, with a 4-6 week taper begun as symptoms start to resolve.

“In a patient who has profound hypoxia and shortness of breath, we may want to go higher: We might give them 2 mg/kg twice a day of [methylprednisolone] in the hospital, wait until they get better, and then slowly taper them, whether it be over 4 or 6 weeks or longer. Occasionally our patients need to get even higher doses of steroids and there’s really nothing to guide us. Who knows if 1 gram of [methylprednisolone] may be better than 60 g of prednisone? But we do it, and patients do get better with the higher doses,” he said.

In rare instances, patients may required other immunosuppressive agents, such as infliximab (Remicade), cyclophosphamide, or mycophenolate mofetil.

Patients who require additional immunosuppressive agents to resolve severe pneumonitis tend to have poor outcomes, Dr. Gettinger said.

Re-challenge of patients with a PD-1 or PD-L1 inhibitor following resolution of pneumonitis appears to be inadvisable for all patients except possibly those with grade 1 (asymptomatic) toxicity, due to the high risk of recurrence, he said.

BOSTON – Treatment of pneumonitis associated with the PD-1 axis checkpoint inhibitors requires close monitoring of patients and rapid clinical response, an oncologist advised.

“We need to be vigilant when we treat our patients. If a patient has cough or shortness of breath, you take it seriously, even if [it is] a patient who has lung cancer and is a smoker,” said Dr. Scott Gettinger of Yale Cancer Center, New Haven, Conn.

He recommended that before starting patients on a programmed death-1 (PD-1) axis inhibitor such as pembrolizumab (Keytruda) or nivolumab (Opdivo), clinicians get baseline oxygen saturation levels to obtain an objective measure for following patients during therapy.

“When you suspect pneumonitis you have to start steroids right away, or patients can spiral down,” he said at the AACR/NCI/EORTC International Conference on Molecular Targets and Cancer Therapeutics.

Pneumonitis – characterized by cough, dyspnea, and hypoxia – is a common adverse event associated with PD-1 and PD-ligand 1 (PD-L1) inhibitors, but is rarely seen in patients treated with CTLA-4 inhibitors such as ipilimumab (Yervoy). Current evidence suggests that pneumonitis is more prevalent in patients with non–small-cell lung cancer, occurring in approximately 4%-6% of patients cases, than in melanoma (1%). The difference is probably due to a history of smoking common to the majority of patients with NSCLC, Dr. Gettinger said.

“The other theme that we’re beginning to see is that maybe pneumonitis is a bit more common with anti-PD-1 vs. anti-PD-L1 antibodies,” he said.

It’s theorized that PD-1 inhibitors may block binding of PD-L2 to one of its binding partners, thereby interfering with respiratory tolerance, he said.

Evidence from the pivotal clinical trial of nivolumab in lung cancer (Checkmate 057) suggests that the time to onset of pneumonitis was a median of 31.1 weeks (range 11.7 to 56.9 weeks). The pneumonitis resolved in about 5-7 weeks.

Investigators at Memorial Sloan Kettering Cancer Center in New York City reported at the European Cancer Congress 2015 on pneumonitis in 36 of 653 patients treated with an anti PD-1/PD-L1 monoclonal antibody from 2009 through 2014, 33 of whom had received a PD-1 inhibitor, and 3 of whom received a PD-L1 inhibitor.

They found that pneumonitis in patients with lung cancer tended to resemble chronic obstructive pneumonia, whereas patients with melanoma were more likely to have ground-glass opacifications on radiography. There was a trend, falling just short of significance, toward association of COP-like pneumonitis with development of grade 3 or greater, and with a requirement for more than one type of immunosuppression, compared to other subtypes.

Algorithms for management

Dr. Gettinger briefly outlined an algorithm offered by Bristol-Myers Squibb for management of suspected pulmonary toxicity with nivolumab. For management of patients with grade 1 toxicities (asymptomatic, with radiographic changes only), it recommends that clinicians consider delay of immuno-oncologic (I-O) therapy, monitor for symptoms every 2-3 days, and consider consultations with pulmonary and infectious disease specialists.

For grade 2 pneumonitis, marked by mild to moderate new symptoms, the algorithm calls for clinicians to delay I-O therapy per protocol, consult with pulmonary and ID specialists, monitor symptoms daily and consider hospitalization, start the patient on steroids with 1.0 mg/kg per day methylprednisolone IV or the oral equivalent, and consider bronchoscopy and lung biopsy.

For patients with grade 3 or 4 toxicities, marked by severe new symptoms, new or worsening hypoxia, or other life-threatening symptoms, the algorithm states that clinicians should discontinue I-O therapy, hospitalize the patient, consult with pulmonary and ID, give 2-4 mg/kg per day methylprednisolone IV or the oral equivalent, add prophylactic antibiotics for opportunistic infections, and consider bronchoscopy and lung biopsy.

At Yale, Dr. Gettinger and colleagues, when presented with a symptomatic patient on a PD-axis inhibitor, will first rule out other etiologies such as infection, chronic obstructive pulmonary disease exacerbation, or cancer progression), typically with bronchoscopy.

For patients with moderate pneumonitis, they may treat with prednisone for 1 or 2 weeks, with a 4-6 week taper begun as symptoms start to resolve.

“In a patient who has profound hypoxia and shortness of breath, we may want to go higher: We might give them 2 mg/kg twice a day of [methylprednisolone] in the hospital, wait until they get better, and then slowly taper them, whether it be over 4 or 6 weeks or longer. Occasionally our patients need to get even higher doses of steroids and there’s really nothing to guide us. Who knows if 1 gram of [methylprednisolone] may be better than 60 g of prednisone? But we do it, and patients do get better with the higher doses,” he said.

In rare instances, patients may required other immunosuppressive agents, such as infliximab (Remicade), cyclophosphamide, or mycophenolate mofetil.

Patients who require additional immunosuppressive agents to resolve severe pneumonitis tend to have poor outcomes, Dr. Gettinger said.

Re-challenge of patients with a PD-1 or PD-L1 inhibitor following resolution of pneumonitis appears to be inadvisable for all patients except possibly those with grade 1 (asymptomatic) toxicity, due to the high risk of recurrence, he said.

BOSTON – An antibody conjugated to the active metabolite of irinotecan was associated with a good overall response rate in patients with heavily pretreated triple-negative breast cancer, with lower toxicities than seen with systemic irinotecan therapy, investigators reported.

Among 54 patients with triple-negative breast cancer (TNBC) in a phase II trial who had undergone a median of six prior lines of therapy, treatment with the conjugate, labeled IMMU-132 (sacituzumab govitecan) was associated with a 31.5% overall response rate, including two complete responses, reported Dr. Aditya Bardia of Massachusetts General Hospital, Boston, and his colleagues, at the AACR–NCI–EORTC International Conference on Molecular Targets and Cancer Therapeutics.

The compound consists of a monoclonal antibody (RS7-3G11) targeted to the Trop-2/EGP-1 panepithelial cancer antigen that is conjugated with the active metabolite of irinotecan (SN-38). The antibody has been shown in preclinical studies to bind to human breast, lung, colon, renal, prostate, and urothelial cancer and other solid malignancies. Trop-2 is expressed in more than 80% of triple-negative breast cancers, said coauthor Dr. David M. Goldenberg, chairman of Immunomedics, the maker of the compound.

“What distinguishes this antibody-drug conjugate, or ADC, is that we have used the active metabolite of irinotecan, which is 100 to 1,000 times more toxic than its parent drug, and the reason it’s tolerated is that it’s conjugated to the antibody, and we have shown preclinically that we can deliver approximately 136 times more SN-38 to the cancer cell than if you give irinotecan and measure how much SN-38 gets to the tumor,” he said at a briefing.

The conjugate has good activity in patients who have experienced relapses after multiple prior lines of therapy, and it can be delivered repeatedly over a long course of therapy with toxicities that are manageable, he noted.

The compound was found to have good activity in several solid tumors in a phase I trial, which led to an expanded phase II trial including, as of May 10, 2015, a total of 56 patients with TNBC, 2 of whom had received fewer than three doses of the study drug by the data cutoff, and therefore were not included in the efficacy analysis.

Patients received IMMU-132 intravenously in doses of 8 or 10 mg/kg on days 1 and 8 of each 21-day cycle, which could be repeated until progression or unacceptable toxicity.

Among 54 assessable patients, the overall response rate (percentage change from baseline according to RECIST [Response Evaluation Criteria in Solid Tumors] 1.1 guidelines) was 31.5%, (17 of 54) consisting of 2 confirmed complete responses and 15 partial responses. Additionally, there were 24 cases of stable disease (4 confirmed), 9 of which had a greater than 20% regression of tumor, but had not met the definition of a partial response by the time of the analysis.

The combined clinical benefit rate, a composite of complete and partial responses and stable disease for 6 months or more, was 44%.

Median progression-free survival in an intent to treat analysis was 7 months. Median overall survival had not been reached, with 87% of patients alive at the data cutoff.

Among all patients in the study, including those with TNBC and other tumors, adverse events at the 10 mg/kg dose included diarrhea in 37% (grade 3 or 4 in 6%); Dr. Goldenberg noted that severe diarrhea is one of the common adverse effects of the parent compound irinotecan, earning it a black-box warning. Other common adverse events associated with irinotecan were neutropenia in 26% (grade 3 or 4 in 15%), febrile neutropenia, all grade 3 or 4, in 4%, and anemia in 20% (grade 3/4 in 6%).

“I have to say that for me, SN-38 was a brilliant idea,” commented Dr. Lee J. Helman, a sarcoma specialist at the National Cancer Institute in Bethesda, Md.

He said that the data showing less severe diarrhea with IMMU-132 than with the parent irinotecan are encouraging, because it may allow more patients to benefit from the therapy.

“This is a very active drug, irinotecan. It gets activated to SN-38 through enterohepatic metabolism, and therefore, it has a very difficult toxicity, which is diarrhea. In fact, I’ve had patients who are responding, who have come off the drug because they can’t stand the diarrhea,” he said.

Dr. Helman moderated a briefing in which Dr. Goldenberg presented the data but was not involved in the study.

BOSTON – An antibody conjugated to the active metabolite of irinotecan was associated with a good overall response rate in patients with heavily pretreated triple-negative breast cancer, with lower toxicities than seen with systemic irinotecan therapy, investigators reported.

Among 54 patients with triple-negative breast cancer (TNBC) in a phase II trial who had undergone a median of six prior lines of therapy, treatment with the conjugate, labeled IMMU-132 (sacituzumab govitecan) was associated with a 31.5% overall response rate, including two complete responses, reported Dr. Aditya Bardia of Massachusetts General Hospital, Boston, and his colleagues, at the AACR–NCI–EORTC International Conference on Molecular Targets and Cancer Therapeutics.

The compound consists of a monoclonal antibody (RS7-3G11) targeted to the Trop-2/EGP-1 panepithelial cancer antigen that is conjugated with the active metabolite of irinotecan (SN-38). The antibody has been shown in preclinical studies to bind to human breast, lung, colon, renal, prostate, and urothelial cancer and other solid malignancies. Trop-2 is expressed in more than 80% of triple-negative breast cancers, said coauthor Dr. David M. Goldenberg, chairman of Immunomedics, the maker of the compound.

“What distinguishes this antibody-drug conjugate, or ADC, is that we have used the active metabolite of irinotecan, which is 100 to 1,000 times more toxic than its parent drug, and the reason it’s tolerated is that it’s conjugated to the antibody, and we have shown preclinically that we can deliver approximately 136 times more SN-38 to the cancer cell than if you give irinotecan and measure how much SN-38 gets to the tumor,” he said at a briefing.

The conjugate has good activity in patients who have experienced relapses after multiple prior lines of therapy, and it can be delivered repeatedly over a long course of therapy with toxicities that are manageable, he noted.

The compound was found to have good activity in several solid tumors in a phase I trial, which led to an expanded phase II trial including, as of May 10, 2015, a total of 56 patients with TNBC, 2 of whom had received fewer than three doses of the study drug by the data cutoff, and therefore were not included in the efficacy analysis.

Patients received IMMU-132 intravenously in doses of 8 or 10 mg/kg on days 1 and 8 of each 21-day cycle, which could be repeated until progression or unacceptable toxicity.

Among 54 assessable patients, the overall response rate (percentage change from baseline according to RECIST [Response Evaluation Criteria in Solid Tumors] 1.1 guidelines) was 31.5%, (17 of 54) consisting of 2 confirmed complete responses and 15 partial responses. Additionally, there were 24 cases of stable disease (4 confirmed), 9 of which had a greater than 20% regression of tumor, but had not met the definition of a partial response by the time of the analysis.

The combined clinical benefit rate, a composite of complete and partial responses and stable disease for 6 months or more, was 44%.

Median progression-free survival in an intent to treat analysis was 7 months. Median overall survival had not been reached, with 87% of patients alive at the data cutoff.

Among all patients in the study, including those with TNBC and other tumors, adverse events at the 10 mg/kg dose included diarrhea in 37% (grade 3 or 4 in 6%); Dr. Goldenberg noted that severe diarrhea is one of the common adverse effects of the parent compound irinotecan, earning it a black-box warning. Other common adverse events associated with irinotecan were neutropenia in 26% (grade 3 or 4 in 15%), febrile neutropenia, all grade 3 or 4, in 4%, and anemia in 20% (grade 3/4 in 6%).

“I have to say that for me, SN-38 was a brilliant idea,” commented Dr. Lee J. Helman, a sarcoma specialist at the National Cancer Institute in Bethesda, Md.

He said that the data showing less severe diarrhea with IMMU-132 than with the parent irinotecan are encouraging, because it may allow more patients to benefit from the therapy.

“This is a very active drug, irinotecan. It gets activated to SN-38 through enterohepatic metabolism, and therefore, it has a very difficult toxicity, which is diarrhea. In fact, I’ve had patients who are responding, who have come off the drug because they can’t stand the diarrhea,” he said.

Dr. Helman moderated a briefing in which Dr. Goldenberg presented the data but was not involved in the study.

BOSTON – An antibody conjugated to the active metabolite of irinotecan was associated with a good overall response rate in patients with heavily pretreated triple-negative breast cancer, with lower toxicities than seen with systemic irinotecan therapy, investigators reported.

Among 54 patients with triple-negative breast cancer (TNBC) in a phase II trial who had undergone a median of six prior lines of therapy, treatment with the conjugate, labeled IMMU-132 (sacituzumab govitecan) was associated with a 31.5% overall response rate, including two complete responses, reported Dr. Aditya Bardia of Massachusetts General Hospital, Boston, and his colleagues, at the AACR–NCI–EORTC International Conference on Molecular Targets and Cancer Therapeutics.

The compound consists of a monoclonal antibody (RS7-3G11) targeted to the Trop-2/EGP-1 panepithelial cancer antigen that is conjugated with the active metabolite of irinotecan (SN-38). The antibody has been shown in preclinical studies to bind to human breast, lung, colon, renal, prostate, and urothelial cancer and other solid malignancies. Trop-2 is expressed in more than 80% of triple-negative breast cancers, said coauthor Dr. David M. Goldenberg, chairman of Immunomedics, the maker of the compound.

“What distinguishes this antibody-drug conjugate, or ADC, is that we have used the active metabolite of irinotecan, which is 100 to 1,000 times more toxic than its parent drug, and the reason it’s tolerated is that it’s conjugated to the antibody, and we have shown preclinically that we can deliver approximately 136 times more SN-38 to the cancer cell than if you give irinotecan and measure how much SN-38 gets to the tumor,” he said at a briefing.

The conjugate has good activity in patients who have experienced relapses after multiple prior lines of therapy, and it can be delivered repeatedly over a long course of therapy with toxicities that are manageable, he noted.

The compound was found to have good activity in several solid tumors in a phase I trial, which led to an expanded phase II trial including, as of May 10, 2015, a total of 56 patients with TNBC, 2 of whom had received fewer than three doses of the study drug by the data cutoff, and therefore were not included in the efficacy analysis.

Patients received IMMU-132 intravenously in doses of 8 or 10 mg/kg on days 1 and 8 of each 21-day cycle, which could be repeated until progression or unacceptable toxicity.

Among 54 assessable patients, the overall response rate (percentage change from baseline according to RECIST [Response Evaluation Criteria in Solid Tumors] 1.1 guidelines) was 31.5%, (17 of 54) consisting of 2 confirmed complete responses and 15 partial responses. Additionally, there were 24 cases of stable disease (4 confirmed), 9 of which had a greater than 20% regression of tumor, but had not met the definition of a partial response by the time of the analysis.

The combined clinical benefit rate, a composite of complete and partial responses and stable disease for 6 months or more, was 44%.

Median progression-free survival in an intent to treat analysis was 7 months. Median overall survival had not been reached, with 87% of patients alive at the data cutoff.

Among all patients in the study, including those with TNBC and other tumors, adverse events at the 10 mg/kg dose included diarrhea in 37% (grade 3 or 4 in 6%); Dr. Goldenberg noted that severe diarrhea is one of the common adverse effects of the parent compound irinotecan, earning it a black-box warning. Other common adverse events associated with irinotecan were neutropenia in 26% (grade 3 or 4 in 15%), febrile neutropenia, all grade 3 or 4, in 4%, and anemia in 20% (grade 3/4 in 6%).

“I have to say that for me, SN-38 was a brilliant idea,” commented Dr. Lee J. Helman, a sarcoma specialist at the National Cancer Institute in Bethesda, Md.

He said that the data showing less severe diarrhea with IMMU-132 than with the parent irinotecan are encouraging, because it may allow more patients to benefit from the therapy.

“This is a very active drug, irinotecan. It gets activated to SN-38 through enterohepatic metabolism, and therefore, it has a very difficult toxicity, which is diarrhea. In fact, I’ve had patients who are responding, who have come off the drug because they can’t stand the diarrhea,” he said.

Dr. Helman moderated a briefing in which Dr. Goldenberg presented the data but was not involved in the study.

BOSTON– Immune-related adverse events associated with checkpoint inhibitor therapy are generally mild to moderate and transient, but some less common side effects can be serious or even fatal, according to an immunotherapy researcher.

“Rapid identification of these side effects and initiation of systemic immunosuppression can improve outcomes without compromising the efficacy of immune-checkpoint inhibition,” said Dr. Antoine Italiano from the Institut Bergonié in Bordeaux, France.

There is also evidence to suggest that immune-related adverse events (irAEs) associated with the programmed-death 1 (PD-1) inhibitors pembrolizumab (Keytruda) and nivolumab (Opdivo) may be predictive of favorable outcomes. In contrast, although there was early clinical evidence to suggest that adverse reactions to immune checkpoint inhibition with cytotoxic T-lymphocyte–associated protein 4 (anti-CTLA-4) antibodies such as ipilimumab (Yervoy) correlate with outcomes, more recent evidence suggests that toxicity with this class of agents is not predictive of efficacy, Dr. Italiano said at the AACR–NCI–EORTC International Conference on Molecular Targets and Cancer Therapeutics.

“Correlation between safety profile and outcome must be confirmed by further studies,” he said, adding that “further studies are also needed to identify patients at high risk of poor tolerability.”

Immune-related adverse events associated with anti-CTLA-4 therapy generally involve organ systems such as the skin, digestive tract, and endocrine system. Rare adverse events reported with these agents include renal injury, sarcoidosis, uveitis, and myelitis, among others.

The events tend to arise around 10 weeks of therapy, following three cycles with either ipilimumab or the investigational agent tremelimumab. Late-occurring events, defined as those that arise more than 70 days after the last infusion, are uncommon, occurring in less than 7% of patients.

Most irAEs seen with anti-CTLA-4 therapy are reversible within about 6 weeks, although some events, such as hypophysitis (autoimmune inflammation of the pituitary gland), can take significantly longer to resolve, Dr. Italiano said.

He cited a recent systematic review and meta-analysis showing that among patients treated with any anti-CTLA-4, the overall incidence of all-grade irAEs was 72%, and the overall incidence of high-grade irAEs was 24%. This study also showed that there was a dose-dependent risk of developing irAEs with ipilimumab, with the incidence of all grades of events at 61% for the 3 mg/kg dose, and 79% for the 10 mg/kg dose.

Two potential biomarkers for gastrointestinal irAEs, the neutrophil-activation markers CD177 and CEACAM1, were identified in a 2013 study. This finding suggests a possible role of neutrophils in ipilimumab-associated GI irAEs, Dr. Italiano noted.

Evidence from early clinical studies of ipilimumab in metastatic melanoma suggested that irAEs correlated with outcomes, but a study published in October 2015 seems to debunk this notion, showing that among 298 patients treated with ipilimumab, neither time to treatment failure nor overall survival were affected by the occurrence of irAEs, he added.

As to whether therapy with anti-CTLA-4 antibodies is safe for treatment of cancer for patients with autoimmune diseases or immunodeficient states, the jury is still out, because these patients were typically excluded from clinical trials.

“But there are a few recent case reports suggesting that treating patients with autoimmune disease with ipilimumab is safe and does not induce exacerbation of the symptoms of the underlying autoimmune disease,” Dr. Italiano said.

PD-1 inhibitors

Adverse events common to the PD-1 inhibitors pembrolizumab and nivolumab and occurring in more than 5% of patients with each include fatigue/asthenia, decreased appetite, diarrhea, rash, pruritus, nausea, and arthralgia.

In clinical trials of the agents for treatment of melanoma, vitiligo was the most common irAE, occurring in 7%-8% of patients. Other events, occurring in similar frequency across the various trials, included hypo- or hyperthyroidism, pneumonitis, colitis, hepatitis, renal failure/nephritis, uveitis/iritis, and hypophysitis.

The time to first occurrence and resolution of irAEs with the PD-1 inhibitors varies by organ system, with skin toxicity occurring within the few weeks of therapy, peaking at about 15 weeks, and resolving by about 25 weeks. Gastrointestinal toxicities crop up at about 10 weeks, but quickly resolve.

Among the less common (less than 10%) irAEs, hepatic and pulmonary events seen to occur around week 8 or 9 and resolve within 2-4 weeks, whereas endocrine events start showing up around week 10, peak at about 25 weeks, and resolve around 40 weeks.

Among patients treated with PD-1 inhibitors for non–small cell lung cancer, the adverse-event profile is similar to that seen in treatment of patients with melanoma, except for the absence of vitiligo, Dr. Italiano noted.

In contrast to the CTLA-4 inhibitors, irAEs seen with the PD-1 inhibitors, especially cutaneous events, appear to be associated with favorable outcomes.

For example, in a prospective, single-center observational study of pembrolizumab in 67 patients with metastatic melanoma, 17 developed vitiligo, and 12 of these patients had an objective response (18% complete and 53% partial responses). The objective response rate in this group was 71%, compared with 28% (14 of 50 patients) for those who did not develop vitiligo.

In a second, retrospective study of 83 patients enrolled in clinical trials of pembrolizumab for melanoma, non–small cell lung cancer, prostate cancer, and Merkel cell carcinoma, patients in each of three pembrolizumab dosing groups who developed cutaneous AEs had significantly longer progression-free intervals than patients who did not develop cutaneous AEs.

A similar correlation between cutaneous events with nivolumab and favorable outcomes was seen in a study of pooled data on 148 patients with resected or unresectable metastatic melanoma. The investigators found that both rash and vitiligo correlated significantly with better overall survival.

BOSTON– Immune-related adverse events associated with checkpoint inhibitor therapy are generally mild to moderate and transient, but some less common side effects can be serious or even fatal, according to an immunotherapy researcher.

“Rapid identification of these side effects and initiation of systemic immunosuppression can improve outcomes without compromising the efficacy of immune-checkpoint inhibition,” said Dr. Antoine Italiano from the Institut Bergonié in Bordeaux, France.

There is also evidence to suggest that immune-related adverse events (irAEs) associated with the programmed-death 1 (PD-1) inhibitors pembrolizumab (Keytruda) and nivolumab (Opdivo) may be predictive of favorable outcomes. In contrast, although there was early clinical evidence to suggest that adverse reactions to immune checkpoint inhibition with cytotoxic T-lymphocyte–associated protein 4 (anti-CTLA-4) antibodies such as ipilimumab (Yervoy) correlate with outcomes, more recent evidence suggests that toxicity with this class of agents is not predictive of efficacy, Dr. Italiano said at the AACR–NCI–EORTC International Conference on Molecular Targets and Cancer Therapeutics.

“Correlation between safety profile and outcome must be confirmed by further studies,” he said, adding that “further studies are also needed to identify patients at high risk of poor tolerability.”

Immune-related adverse events associated with anti-CTLA-4 therapy generally involve organ systems such as the skin, digestive tract, and endocrine system. Rare adverse events reported with these agents include renal injury, sarcoidosis, uveitis, and myelitis, among others.

The events tend to arise around 10 weeks of therapy, following three cycles with either ipilimumab or the investigational agent tremelimumab. Late-occurring events, defined as those that arise more than 70 days after the last infusion, are uncommon, occurring in less than 7% of patients.

Most irAEs seen with anti-CTLA-4 therapy are reversible within about 6 weeks, although some events, such as hypophysitis (autoimmune inflammation of the pituitary gland), can take significantly longer to resolve, Dr. Italiano said.

He cited a recent systematic review and meta-analysis showing that among patients treated with any anti-CTLA-4, the overall incidence of all-grade irAEs was 72%, and the overall incidence of high-grade irAEs was 24%. This study also showed that there was a dose-dependent risk of developing irAEs with ipilimumab, with the incidence of all grades of events at 61% for the 3 mg/kg dose, and 79% for the 10 mg/kg dose.

Two potential biomarkers for gastrointestinal irAEs, the neutrophil-activation markers CD177 and CEACAM1, were identified in a 2013 study. This finding suggests a possible role of neutrophils in ipilimumab-associated GI irAEs, Dr. Italiano noted.

Evidence from early clinical studies of ipilimumab in metastatic melanoma suggested that irAEs correlated with outcomes, but a study published in October 2015 seems to debunk this notion, showing that among 298 patients treated with ipilimumab, neither time to treatment failure nor overall survival were affected by the occurrence of irAEs, he added.

As to whether therapy with anti-CTLA-4 antibodies is safe for treatment of cancer for patients with autoimmune diseases or immunodeficient states, the jury is still out, because these patients were typically excluded from clinical trials.

“But there are a few recent case reports suggesting that treating patients with autoimmune disease with ipilimumab is safe and does not induce exacerbation of the symptoms of the underlying autoimmune disease,” Dr. Italiano said.

PD-1 inhibitors

Adverse events common to the PD-1 inhibitors pembrolizumab and nivolumab and occurring in more than 5% of patients with each include fatigue/asthenia, decreased appetite, diarrhea, rash, pruritus, nausea, and arthralgia.

In clinical trials of the agents for treatment of melanoma, vitiligo was the most common irAE, occurring in 7%-8% of patients. Other events, occurring in similar frequency across the various trials, included hypo- or hyperthyroidism, pneumonitis, colitis, hepatitis, renal failure/nephritis, uveitis/iritis, and hypophysitis.

The time to first occurrence and resolution of irAEs with the PD-1 inhibitors varies by organ system, with skin toxicity occurring within the few weeks of therapy, peaking at about 15 weeks, and resolving by about 25 weeks. Gastrointestinal toxicities crop up at about 10 weeks, but quickly resolve.

Among the less common (less than 10%) irAEs, hepatic and pulmonary events seen to occur around week 8 or 9 and resolve within 2-4 weeks, whereas endocrine events start showing up around week 10, peak at about 25 weeks, and resolve around 40 weeks.

Among patients treated with PD-1 inhibitors for non–small cell lung cancer, the adverse-event profile is similar to that seen in treatment of patients with melanoma, except for the absence of vitiligo, Dr. Italiano noted.

In contrast to the CTLA-4 inhibitors, irAEs seen with the PD-1 inhibitors, especially cutaneous events, appear to be associated with favorable outcomes.

For example, in a prospective, single-center observational study of pembrolizumab in 67 patients with metastatic melanoma, 17 developed vitiligo, and 12 of these patients had an objective response (18% complete and 53% partial responses). The objective response rate in this group was 71%, compared with 28% (14 of 50 patients) for those who did not develop vitiligo.

In a second, retrospective study of 83 patients enrolled in clinical trials of pembrolizumab for melanoma, non–small cell lung cancer, prostate cancer, and Merkel cell carcinoma, patients in each of three pembrolizumab dosing groups who developed cutaneous AEs had significantly longer progression-free intervals than patients who did not develop cutaneous AEs.

A similar correlation between cutaneous events with nivolumab and favorable outcomes was seen in a study of pooled data on 148 patients with resected or unresectable metastatic melanoma. The investigators found that both rash and vitiligo correlated significantly with better overall survival.

BOSTON– Immune-related adverse events associated with checkpoint inhibitor therapy are generally mild to moderate and transient, but some less common side effects can be serious or even fatal, according to an immunotherapy researcher.

“Rapid identification of these side effects and initiation of systemic immunosuppression can improve outcomes without compromising the efficacy of immune-checkpoint inhibition,” said Dr. Antoine Italiano from the Institut Bergonié in Bordeaux, France.

There is also evidence to suggest that immune-related adverse events (irAEs) associated with the programmed-death 1 (PD-1) inhibitors pembrolizumab (Keytruda) and nivolumab (Opdivo) may be predictive of favorable outcomes. In contrast, although there was early clinical evidence to suggest that adverse reactions to immune checkpoint inhibition with cytotoxic T-lymphocyte–associated protein 4 (anti-CTLA-4) antibodies such as ipilimumab (Yervoy) correlate with outcomes, more recent evidence suggests that toxicity with this class of agents is not predictive of efficacy, Dr. Italiano said at the AACR–NCI–EORTC International Conference on Molecular Targets and Cancer Therapeutics.

“Correlation between safety profile and outcome must be confirmed by further studies,” he said, adding that “further studies are also needed to identify patients at high risk of poor tolerability.”

Immune-related adverse events associated with anti-CTLA-4 therapy generally involve organ systems such as the skin, digestive tract, and endocrine system. Rare adverse events reported with these agents include renal injury, sarcoidosis, uveitis, and myelitis, among others.

The events tend to arise around 10 weeks of therapy, following three cycles with either ipilimumab or the investigational agent tremelimumab. Late-occurring events, defined as those that arise more than 70 days after the last infusion, are uncommon, occurring in less than 7% of patients.

Most irAEs seen with anti-CTLA-4 therapy are reversible within about 6 weeks, although some events, such as hypophysitis (autoimmune inflammation of the pituitary gland), can take significantly longer to resolve, Dr. Italiano said.

He cited a recent systematic review and meta-analysis showing that among patients treated with any anti-CTLA-4, the overall incidence of all-grade irAEs was 72%, and the overall incidence of high-grade irAEs was 24%. This study also showed that there was a dose-dependent risk of developing irAEs with ipilimumab, with the incidence of all grades of events at 61% for the 3 mg/kg dose, and 79% for the 10 mg/kg dose.

Two potential biomarkers for gastrointestinal irAEs, the neutrophil-activation markers CD177 and CEACAM1, were identified in a 2013 study. This finding suggests a possible role of neutrophils in ipilimumab-associated GI irAEs, Dr. Italiano noted.

Evidence from early clinical studies of ipilimumab in metastatic melanoma suggested that irAEs correlated with outcomes, but a study published in October 2015 seems to debunk this notion, showing that among 298 patients treated with ipilimumab, neither time to treatment failure nor overall survival were affected by the occurrence of irAEs, he added.

As to whether therapy with anti-CTLA-4 antibodies is safe for treatment of cancer for patients with autoimmune diseases or immunodeficient states, the jury is still out, because these patients were typically excluded from clinical trials.

“But there are a few recent case reports suggesting that treating patients with autoimmune disease with ipilimumab is safe and does not induce exacerbation of the symptoms of the underlying autoimmune disease,” Dr. Italiano said.

PD-1 inhibitors

Adverse events common to the PD-1 inhibitors pembrolizumab and nivolumab and occurring in more than 5% of patients with each include fatigue/asthenia, decreased appetite, diarrhea, rash, pruritus, nausea, and arthralgia.

In clinical trials of the agents for treatment of melanoma, vitiligo was the most common irAE, occurring in 7%-8% of patients. Other events, occurring in similar frequency across the various trials, included hypo- or hyperthyroidism, pneumonitis, colitis, hepatitis, renal failure/nephritis, uveitis/iritis, and hypophysitis.

The time to first occurrence and resolution of irAEs with the PD-1 inhibitors varies by organ system, with skin toxicity occurring within the few weeks of therapy, peaking at about 15 weeks, and resolving by about 25 weeks. Gastrointestinal toxicities crop up at about 10 weeks, but quickly resolve.

Among the less common (less than 10%) irAEs, hepatic and pulmonary events seen to occur around week 8 or 9 and resolve within 2-4 weeks, whereas endocrine events start showing up around week 10, peak at about 25 weeks, and resolve around 40 weeks.

Among patients treated with PD-1 inhibitors for non–small cell lung cancer, the adverse-event profile is similar to that seen in treatment of patients with melanoma, except for the absence of vitiligo, Dr. Italiano noted.

In contrast to the CTLA-4 inhibitors, irAEs seen with the PD-1 inhibitors, especially cutaneous events, appear to be associated with favorable outcomes.

For example, in a prospective, single-center observational study of pembrolizumab in 67 patients with metastatic melanoma, 17 developed vitiligo, and 12 of these patients had an objective response (18% complete and 53% partial responses). The objective response rate in this group was 71%, compared with 28% (14 of 50 patients) for those who did not develop vitiligo.

In a second, retrospective study of 83 patients enrolled in clinical trials of pembrolizumab for melanoma, non–small cell lung cancer, prostate cancer, and Merkel cell carcinoma, patients in each of three pembrolizumab dosing groups who developed cutaneous AEs had significantly longer progression-free intervals than patients who did not develop cutaneous AEs.

A similar correlation between cutaneous events with nivolumab and favorable outcomes was seen in a study of pooled data on 148 patients with resected or unresectable metastatic melanoma. The investigators found that both rash and vitiligo correlated significantly with better overall survival.

BOSTON – A novel pan-AKT inhibitor was associated in a phase 1 trial with tumor shrinkage in a large proportion of patients with solid tumors bearing a mutation in AKT1.

The investigational compound, labeled AZD5363, targets the AKT1 E17K mutation, which occurs in approximately 0.5% to 4% of many different types of solid tumors. In a phase I trial, 33 of 41 assessable patients with tumors bearing the mutation had regression of tumors, reported Dr. David M. Hyman, acting director of experimental therapeutics at Memorial Sloan Kettering Cancer Center in New York.

“We believe this is the first clinical data suggesting that the AKT1 E17K mutations are targetable driver mutations in solid tumors,” he said at a briefing at the AACR/NCI/EORTC International Conference on Molecular Targets and Cancer Therapeutics.

Mutations in the PI3K/AKT/mTOR pathway are frequently found in human cancers, and dysregulation of the pathway has been shown to drive tumor growth and survival of malignant cells.

“Although this pathway is among the most commonly activated in all human cancer, it is rarely activated by somatic mutation of AKT itself,” Dr. Hyman said. “When you see AKT mutations, they are almost always of a single allele at the 17 amino acid position, which is the E17K mutation. This is a mutation that has been extensively functionally classified, and behaves as a classic driver oncogene.”

The E17K mutation occurs in approximately 4% of breast cancers, 2% of bladder cancers, 1.5% of colorectal tumors, 1% each of cervical, ovarian, and prostate cancers, and 0.5% of lung adenocarcinomas.

AZD5363 is a catalytic inhibitor of all AKT isoforms 1, 2, and 3. In preclinical studies, it has been shown to inhibit tumor cell proliferation in vitro, and tumor growth in vivo in tumor xenograft models.

Dr. Hyman and colleagues reported on early results from one part of a four-part, phase I study of AZD5363 in adults with advanced solid tumors. Part D of the ongoing study focuses on patients with tumors bearing AKT1 mutations.

At present, 45 patients with E17K mutations have been treated with the agent, including 21 with estrogen receptor (ER)-positive breast cancer, four with triple-negative breast cancer, 14 with gynecological cancers, and eight with other tumor types.

As noted, of 41 patients with sufficient data for assessment, 33 had some degree of tumor regression, and of this group 12 met RECIST (Response Evaluation Criteria in Solid Tumors) parameters for partial responses. Among patients assessable for response, 14 of 18 with ER-positive breast cancer had lesion shrinkage, and five of these patients had partial responses (3 confirmed and 2 unconfirmed).

Among patients with gynecological tumors, 9 of 11 had shrinkage of their target lesion, including 3 with confirmed partial responses. Two of these patients are still on study, with the longest follow-up at the time of data cutoff of 39.2 months

Among patients with 12 other tumor types (apocrine anal, triple-negative breast cancer, colon, lung, prostate, squamous anal, thyroid) 10 patients had demonstrated target lesion shrinkage, one had a confirmed partial response that is ongoing, and two had unconfirmed partial response, one of which is ongoing.

The investigators also used cell-free DNA (cfDNA) analysis to evaluate efficacy of AZD5363 during treatment, and found that declines in AKT1 cfDNA allele fractions were transient among non-responders, whereas declines persisting for 21 days or more correlated with both durable tumor regression and RECIST response (P = .0049).

Dr. Hyman noted that there were also preclinical signs that AZD5363 could lead to reactivation of ER signaling.

“It became a natural question to ask whether patients did better with a combination of ER-directed therapy and AKT-directed therapy,” he said.

To answer this question, the investigators offered sequential therapy to patients with ER-positive breast cancer who had previously had resistance to the ER degrader fulvestrant and who had progressed on AZD5363 monotherapy. Two patients were offered crossover to AZD5363 with fulvestrant, regardless of prior hormonal therapy.

One of the patients had disease progression while on the combination. The second patient, who had received 10 prior lines of therapy and had disease progression before starting on the combination at 120 days on treatment, had significant shrinkage of her breast lesion 6 weeks after starting on the combination, and liver lesions that had previously enhanced on imaging are no longer enhancing, and her cfDNA allele fractions have declined. She remains on the combination after more than 160 days.

“This was very convincing evidence to us of synergy between these two compounds,” he said.

Dr. Jean-Charles Soria, from the Institut Gustave Roussy, Villejuif, France, who moderated the briefing, commented that “although this [mutation] is a rare situation, in the range of 2% to 3%, it’s certainly important, because 2% to 3% of breast cancer patients is a significant number, so this drug has legs,” he said.

The question yet to be settled, however, is the true duration of response, he added.

The study is sponsored by AstraZeneca. Dr. Hyman disclosed consulting for Atara Biotherapeutics. Dr. Soria reported having no disclosures.

BOSTON – A novel pan-AKT inhibitor was associated in a phase 1 trial with tumor shrinkage in a large proportion of patients with solid tumors bearing a mutation in AKT1.

The investigational compound, labeled AZD5363, targets the AKT1 E17K mutation, which occurs in approximately 0.5% to 4% of many different types of solid tumors. In a phase I trial, 33 of 41 assessable patients with tumors bearing the mutation had regression of tumors, reported Dr. David M. Hyman, acting director of experimental therapeutics at Memorial Sloan Kettering Cancer Center in New York.

“We believe this is the first clinical data suggesting that the AKT1 E17K mutations are targetable driver mutations in solid tumors,” he said at a briefing at the AACR/NCI/EORTC International Conference on Molecular Targets and Cancer Therapeutics.

Mutations in the PI3K/AKT/mTOR pathway are frequently found in human cancers, and dysregulation of the pathway has been shown to drive tumor growth and survival of malignant cells.

“Although this pathway is among the most commonly activated in all human cancer, it is rarely activated by somatic mutation of AKT itself,” Dr. Hyman said. “When you see AKT mutations, they are almost always of a single allele at the 17 amino acid position, which is the E17K mutation. This is a mutation that has been extensively functionally classified, and behaves as a classic driver oncogene.”

The E17K mutation occurs in approximately 4% of breast cancers, 2% of bladder cancers, 1.5% of colorectal tumors, 1% each of cervical, ovarian, and prostate cancers, and 0.5% of lung adenocarcinomas.

AZD5363 is a catalytic inhibitor of all AKT isoforms 1, 2, and 3. In preclinical studies, it has been shown to inhibit tumor cell proliferation in vitro, and tumor growth in vivo in tumor xenograft models.

Dr. Hyman and colleagues reported on early results from one part of a four-part, phase I study of AZD5363 in adults with advanced solid tumors. Part D of the ongoing study focuses on patients with tumors bearing AKT1 mutations.

At present, 45 patients with E17K mutations have been treated with the agent, including 21 with estrogen receptor (ER)-positive breast cancer, four with triple-negative breast cancer, 14 with gynecological cancers, and eight with other tumor types.

As noted, of 41 patients with sufficient data for assessment, 33 had some degree of tumor regression, and of this group 12 met RECIST (Response Evaluation Criteria in Solid Tumors) parameters for partial responses. Among patients assessable for response, 14 of 18 with ER-positive breast cancer had lesion shrinkage, and five of these patients had partial responses (3 confirmed and 2 unconfirmed).

Among patients with gynecological tumors, 9 of 11 had shrinkage of their target lesion, including 3 with confirmed partial responses. Two of these patients are still on study, with the longest follow-up at the time of data cutoff of 39.2 months

Among patients with 12 other tumor types (apocrine anal, triple-negative breast cancer, colon, lung, prostate, squamous anal, thyroid) 10 patients had demonstrated target lesion shrinkage, one had a confirmed partial response that is ongoing, and two had unconfirmed partial response, one of which is ongoing.

The investigators also used cell-free DNA (cfDNA) analysis to evaluate efficacy of AZD5363 during treatment, and found that declines in AKT1 cfDNA allele fractions were transient among non-responders, whereas declines persisting for 21 days or more correlated with both durable tumor regression and RECIST response (P = .0049).

Dr. Hyman noted that there were also preclinical signs that AZD5363 could lead to reactivation of ER signaling.

“It became a natural question to ask whether patients did better with a combination of ER-directed therapy and AKT-directed therapy,” he said.

To answer this question, the investigators offered sequential therapy to patients with ER-positive breast cancer who had previously had resistance to the ER degrader fulvestrant and who had progressed on AZD5363 monotherapy. Two patients were offered crossover to AZD5363 with fulvestrant, regardless of prior hormonal therapy.

One of the patients had disease progression while on the combination. The second patient, who had received 10 prior lines of therapy and had disease progression before starting on the combination at 120 days on treatment, had significant shrinkage of her breast lesion 6 weeks after starting on the combination, and liver lesions that had previously enhanced on imaging are no longer enhancing, and her cfDNA allele fractions have declined. She remains on the combination after more than 160 days.

“This was very convincing evidence to us of synergy between these two compounds,” he said.

Dr. Jean-Charles Soria, from the Institut Gustave Roussy, Villejuif, France, who moderated the briefing, commented that “although this [mutation] is a rare situation, in the range of 2% to 3%, it’s certainly important, because 2% to 3% of breast cancer patients is a significant number, so this drug has legs,” he said.

The question yet to be settled, however, is the true duration of response, he added.

The study is sponsored by AstraZeneca. Dr. Hyman disclosed consulting for Atara Biotherapeutics. Dr. Soria reported having no disclosures.

BOSTON – A novel pan-AKT inhibitor was associated in a phase 1 trial with tumor shrinkage in a large proportion of patients with solid tumors bearing a mutation in AKT1.

The investigational compound, labeled AZD5363, targets the AKT1 E17K mutation, which occurs in approximately 0.5% to 4% of many different types of solid tumors. In a phase I trial, 33 of 41 assessable patients with tumors bearing the mutation had regression of tumors, reported Dr. David M. Hyman, acting director of experimental therapeutics at Memorial Sloan Kettering Cancer Center in New York.

“We believe this is the first clinical data suggesting that the AKT1 E17K mutations are targetable driver mutations in solid tumors,” he said at a briefing at the AACR/NCI/EORTC International Conference on Molecular Targets and Cancer Therapeutics.

Mutations in the PI3K/AKT/mTOR pathway are frequently found in human cancers, and dysregulation of the pathway has been shown to drive tumor growth and survival of malignant cells.

“Although this pathway is among the most commonly activated in all human cancer, it is rarely activated by somatic mutation of AKT itself,” Dr. Hyman said. “When you see AKT mutations, they are almost always of a single allele at the 17 amino acid position, which is the E17K mutation. This is a mutation that has been extensively functionally classified, and behaves as a classic driver oncogene.”

The E17K mutation occurs in approximately 4% of breast cancers, 2% of bladder cancers, 1.5% of colorectal tumors, 1% each of cervical, ovarian, and prostate cancers, and 0.5% of lung adenocarcinomas.

AZD5363 is a catalytic inhibitor of all AKT isoforms 1, 2, and 3. In preclinical studies, it has been shown to inhibit tumor cell proliferation in vitro, and tumor growth in vivo in tumor xenograft models.

Dr. Hyman and colleagues reported on early results from one part of a four-part, phase I study of AZD5363 in adults with advanced solid tumors. Part D of the ongoing study focuses on patients with tumors bearing AKT1 mutations.

At present, 45 patients with E17K mutations have been treated with the agent, including 21 with estrogen receptor (ER)-positive breast cancer, four with triple-negative breast cancer, 14 with gynecological cancers, and eight with other tumor types.

As noted, of 41 patients with sufficient data for assessment, 33 had some degree of tumor regression, and of this group 12 met RECIST (Response Evaluation Criteria in Solid Tumors) parameters for partial responses. Among patients assessable for response, 14 of 18 with ER-positive breast cancer had lesion shrinkage, and five of these patients had partial responses (3 confirmed and 2 unconfirmed).

Among patients with gynecological tumors, 9 of 11 had shrinkage of their target lesion, including 3 with confirmed partial responses. Two of these patients are still on study, with the longest follow-up at the time of data cutoff of 39.2 months

Among patients with 12 other tumor types (apocrine anal, triple-negative breast cancer, colon, lung, prostate, squamous anal, thyroid) 10 patients had demonstrated target lesion shrinkage, one had a confirmed partial response that is ongoing, and two had unconfirmed partial response, one of which is ongoing.

The investigators also used cell-free DNA (cfDNA) analysis to evaluate efficacy of AZD5363 during treatment, and found that declines in AKT1 cfDNA allele fractions were transient among non-responders, whereas declines persisting for 21 days or more correlated with both durable tumor regression and RECIST response (P = .0049).

Dr. Hyman noted that there were also preclinical signs that AZD5363 could lead to reactivation of ER signaling.

“It became a natural question to ask whether patients did better with a combination of ER-directed therapy and AKT-directed therapy,” he said.

To answer this question, the investigators offered sequential therapy to patients with ER-positive breast cancer who had previously had resistance to the ER degrader fulvestrant and who had progressed on AZD5363 monotherapy. Two patients were offered crossover to AZD5363 with fulvestrant, regardless of prior hormonal therapy.

One of the patients had disease progression while on the combination. The second patient, who had received 10 prior lines of therapy and had disease progression before starting on the combination at 120 days on treatment, had significant shrinkage of her breast lesion 6 weeks after starting on the combination, and liver lesions that had previously enhanced on imaging are no longer enhancing, and her cfDNA allele fractions have declined. She remains on the combination after more than 160 days.

“This was very convincing evidence to us of synergy between these two compounds,” he said.

Dr. Jean-Charles Soria, from the Institut Gustave Roussy, Villejuif, France, who moderated the briefing, commented that “although this [mutation] is a rare situation, in the range of 2% to 3%, it’s certainly important, because 2% to 3% of breast cancer patients is a significant number, so this drug has legs,” he said.

The question yet to be settled, however, is the true duration of response, he added.

The study is sponsored by AstraZeneca. Dr. Hyman disclosed consulting for Atara Biotherapeutics. Dr. Soria reported having no disclosures.

BOSTON – An experimental small-molecule tyrosine kinase inhibitor targeted to the epidermal growth factor receptor showed encouraging clinical activity in patients with heavily pretreated non–small cell lung cancer and leptomeningeal disease, one of the most devastating complications of advanced cancer, investigators report.

Of eight patients evaluable for response in a phase I, open label study of the compound, labeled AZD9291, one had confirmed clearance of malignant cells from cerebrospinal fluid (CSF), and four others had suspected but unconfirmed clearance of cancer from CSF, reported Dr. Dan Ho Lee of the University of Ulsan College of Medicine, Seoul, South Korea.

In addition, six patients had investigator-rated improvements in brain imaging, and three of seven with abnormal neurologic exams at baseline had significant improvement of symptoms, Dr. Lee said at the AACR-NCI-EORTC International Conference on Molecular Targets and Cancer Therapeutics.

“Leptomeningeal disease involvement is the nightmare every single clinician taking care of lung cancer patients with EGFR [epidermal growth factor receptor] mutations does not want to face, because that’s a death sentence,” commented Dr. Jean-Charles Soria of the Institut Gustave Roussy, Villejuif, France.

“Having a drug that is delivering in 8 out of 12 patients responding clinically by improvement in their neurological exam, radiologically by MRI, and cytologically by CSF is really transformational for that group of patients,” he said, at a briefing where Dr. Lee presented the data, prior to his presentation in an oral session.

AZD9291 is a tyrosine kinase inhibitor (TKI) selective for EGFR mutations, including T790M resistance mutations, in both preclinical and clinical studies, the compound appeared capable of crossing the blood-brain barrier, making it a candidate for clinical trials in patients with leptomeningeal metastases.

Dr. Lee noted that in patients with EGFR-mutant non–small cell lung cancer (NSCLC), there appears to be an increased risk for CNS involvement, particularly among patients who had received a first-generation EGFR-targeted TKI. Patients with EGFR-mutant NSCLC and leptomeningeal metastases have an expected overall survival of 7-14 months, he said.

As previously reported, in a multicenter clinical trial of AZD9291 (N Engl J Med. 2015 Apr 30. doi: 10.1056/NEJMoa1411817) 123 of 239 patients (51%) who could be evaluated for response had either a partial or complete response. In the subset of patients with known EGFR T790M mutations, the objective response rate was even higher at 61%. In contrast, the objective response rate was 21% in the subset of 61 patients who did not have known EGFR T790M mutations, the investigators said.

Dr. Lee reported here on a cohort of patients with EGFR-mutant NSCLC with leptomeningeal metastases in the BLOOM study. The patients were treated with AZD9291 in a once-daily, 160 mg oral dose.

A total of four men and nine women were enrolled in the cohort. All patients had NSCLC with a sensitizing EGFR mutation, and five were confirmed to have the T790M mutation. The patients had received a median of three prior lines of therapy, including chemotherapy and anti-EGFR TKI. In all, 12 patients were planned for the efficacy analysis, but three withdrew, one at week 6 due to dysphagia, one at week 1 due to aspiration pneumonia, and one at week 3 due to extracranial disease progression.

Of the remaining nine patients, eight had sufficient data to report on all preliminary outcomes, and one had partial data available.

The authors found that, as noted before, three of the eight patients had investigator-rated improvement over baseline in neurologic exams, and of the five patients with normal baseline neurology, four had no change at the most recent follow-up (26 weeks).

Also as noted, one patient had confirmed clearance of cancer from the CSF, and a total of five (including this patient) had negative cytology at their most recent visit. In addition, six of seven patients who had baseline and follow-up brain MRI studies showed improvement at the most recent scan, Dr. Lee said.

The safety profile was similar to that seen with other EGFR-TKIs, with diarrhea and rash, none worse than grade 2, being the most frequent adverse event. There were four grade 3 or greater adverse events, occurring once each in four patients: anemia, hyponatremia, aspiration pneumonia, and neutropenia, the last of which resolved after a 3-day interruption and resumption of the drug at half the original dose (80 mg) daily.

Dr. Soria noted that while the data look very promising, the real test of AZD9291 will come with longer follow-up showing whether the response is durable.

The study was funded by AstraZeneca. Dr. Lee is an advisory board member of Bristol-Myers Squibb, Korea, Boehringer Ingelheim Korea, and Pfizer Korea, and has received lecture fees from AstraZeneca, Ely Lilly, Merck Sharp & Dohme, and Roche. Dr. Soria reported having no disclosures.

BOSTON – An experimental small-molecule tyrosine kinase inhibitor targeted to the epidermal growth factor receptor showed encouraging clinical activity in patients with heavily pretreated non–small cell lung cancer and leptomeningeal disease, one of the most devastating complications of advanced cancer, investigators report.

Of eight patients evaluable for response in a phase I, open label study of the compound, labeled AZD9291, one had confirmed clearance of malignant cells from cerebrospinal fluid (CSF), and four others had suspected but unconfirmed clearance of cancer from CSF, reported Dr. Dan Ho Lee of the University of Ulsan College of Medicine, Seoul, South Korea.

In addition, six patients had investigator-rated improvements in brain imaging, and three of seven with abnormal neurologic exams at baseline had significant improvement of symptoms, Dr. Lee said at the AACR-NCI-EORTC International Conference on Molecular Targets and Cancer Therapeutics.

“Leptomeningeal disease involvement is the nightmare every single clinician taking care of lung cancer patients with EGFR [epidermal growth factor receptor] mutations does not want to face, because that’s a death sentence,” commented Dr. Jean-Charles Soria of the Institut Gustave Roussy, Villejuif, France.

“Having a drug that is delivering in 8 out of 12 patients responding clinically by improvement in their neurological exam, radiologically by MRI, and cytologically by CSF is really transformational for that group of patients,” he said, at a briefing where Dr. Lee presented the data, prior to his presentation in an oral session.

AZD9291 is a tyrosine kinase inhibitor (TKI) selective for EGFR mutations, including T790M resistance mutations, in both preclinical and clinical studies, the compound appeared capable of crossing the blood-brain barrier, making it a candidate for clinical trials in patients with leptomeningeal metastases.

Dr. Lee noted that in patients with EGFR-mutant non–small cell lung cancer (NSCLC), there appears to be an increased risk for CNS involvement, particularly among patients who had received a first-generation EGFR-targeted TKI. Patients with EGFR-mutant NSCLC and leptomeningeal metastases have an expected overall survival of 7-14 months, he said.

As previously reported, in a multicenter clinical trial of AZD9291 (N Engl J Med. 2015 Apr 30. doi: 10.1056/NEJMoa1411817) 123 of 239 patients (51%) who could be evaluated for response had either a partial or complete response. In the subset of patients with known EGFR T790M mutations, the objective response rate was even higher at 61%. In contrast, the objective response rate was 21% in the subset of 61 patients who did not have known EGFR T790M mutations, the investigators said.

Dr. Lee reported here on a cohort of patients with EGFR-mutant NSCLC with leptomeningeal metastases in the BLOOM study. The patients were treated with AZD9291 in a once-daily, 160 mg oral dose.

A total of four men and nine women were enrolled in the cohort. All patients had NSCLC with a sensitizing EGFR mutation, and five were confirmed to have the T790M mutation. The patients had received a median of three prior lines of therapy, including chemotherapy and anti-EGFR TKI. In all, 12 patients were planned for the efficacy analysis, but three withdrew, one at week 6 due to dysphagia, one at week 1 due to aspiration pneumonia, and one at week 3 due to extracranial disease progression.

Of the remaining nine patients, eight had sufficient data to report on all preliminary outcomes, and one had partial data available.

The authors found that, as noted before, three of the eight patients had investigator-rated improvement over baseline in neurologic exams, and of the five patients with normal baseline neurology, four had no change at the most recent follow-up (26 weeks).

Also as noted, one patient had confirmed clearance of cancer from the CSF, and a total of five (including this patient) had negative cytology at their most recent visit. In addition, six of seven patients who had baseline and follow-up brain MRI studies showed improvement at the most recent scan, Dr. Lee said.

The safety profile was similar to that seen with other EGFR-TKIs, with diarrhea and rash, none worse than grade 2, being the most frequent adverse event. There were four grade 3 or greater adverse events, occurring once each in four patients: anemia, hyponatremia, aspiration pneumonia, and neutropenia, the last of which resolved after a 3-day interruption and resumption of the drug at half the original dose (80 mg) daily.

Dr. Soria noted that while the data look very promising, the real test of AZD9291 will come with longer follow-up showing whether the response is durable.

The study was funded by AstraZeneca. Dr. Lee is an advisory board member of Bristol-Myers Squibb, Korea, Boehringer Ingelheim Korea, and Pfizer Korea, and has received lecture fees from AstraZeneca, Ely Lilly, Merck Sharp & Dohme, and Roche. Dr. Soria reported having no disclosures.

BOSTON – An experimental small-molecule tyrosine kinase inhibitor targeted to the epidermal growth factor receptor showed encouraging clinical activity in patients with heavily pretreated non–small cell lung cancer and leptomeningeal disease, one of the most devastating complications of advanced cancer, investigators report.

Of eight patients evaluable for response in a phase I, open label study of the compound, labeled AZD9291, one had confirmed clearance of malignant cells from cerebrospinal fluid (CSF), and four others had suspected but unconfirmed clearance of cancer from CSF, reported Dr. Dan Ho Lee of the University of Ulsan College of Medicine, Seoul, South Korea.

In addition, six patients had investigator-rated improvements in brain imaging, and three of seven with abnormal neurologic exams at baseline had significant improvement of symptoms, Dr. Lee said at the AACR-NCI-EORTC International Conference on Molecular Targets and Cancer Therapeutics.

“Leptomeningeal disease involvement is the nightmare every single clinician taking care of lung cancer patients with EGFR [epidermal growth factor receptor] mutations does not want to face, because that’s a death sentence,” commented Dr. Jean-Charles Soria of the Institut Gustave Roussy, Villejuif, France.

“Having a drug that is delivering in 8 out of 12 patients responding clinically by improvement in their neurological exam, radiologically by MRI, and cytologically by CSF is really transformational for that group of patients,” he said, at a briefing where Dr. Lee presented the data, prior to his presentation in an oral session.

AZD9291 is a tyrosine kinase inhibitor (TKI) selective for EGFR mutations, including T790M resistance mutations, in both preclinical and clinical studies, the compound appeared capable of crossing the blood-brain barrier, making it a candidate for clinical trials in patients with leptomeningeal metastases.

Dr. Lee noted that in patients with EGFR-mutant non–small cell lung cancer (NSCLC), there appears to be an increased risk for CNS involvement, particularly among patients who had received a first-generation EGFR-targeted TKI. Patients with EGFR-mutant NSCLC and leptomeningeal metastases have an expected overall survival of 7-14 months, he said.

As previously reported, in a multicenter clinical trial of AZD9291 (N Engl J Med. 2015 Apr 30. doi: 10.1056/NEJMoa1411817) 123 of 239 patients (51%) who could be evaluated for response had either a partial or complete response. In the subset of patients with known EGFR T790M mutations, the objective response rate was even higher at 61%. In contrast, the objective response rate was 21% in the subset of 61 patients who did not have known EGFR T790M mutations, the investigators said.

Dr. Lee reported here on a cohort of patients with EGFR-mutant NSCLC with leptomeningeal metastases in the BLOOM study. The patients were treated with AZD9291 in a once-daily, 160 mg oral dose.

A total of four men and nine women were enrolled in the cohort. All patients had NSCLC with a sensitizing EGFR mutation, and five were confirmed to have the T790M mutation. The patients had received a median of three prior lines of therapy, including chemotherapy and anti-EGFR TKI. In all, 12 patients were planned for the efficacy analysis, but three withdrew, one at week 6 due to dysphagia, one at week 1 due to aspiration pneumonia, and one at week 3 due to extracranial disease progression.

Of the remaining nine patients, eight had sufficient data to report on all preliminary outcomes, and one had partial data available.

The authors found that, as noted before, three of the eight patients had investigator-rated improvement over baseline in neurologic exams, and of the five patients with normal baseline neurology, four had no change at the most recent follow-up (26 weeks).

Also as noted, one patient had confirmed clearance of cancer from the CSF, and a total of five (including this patient) had negative cytology at their most recent visit. In addition, six of seven patients who had baseline and follow-up brain MRI studies showed improvement at the most recent scan, Dr. Lee said.

The safety profile was similar to that seen with other EGFR-TKIs, with diarrhea and rash, none worse than grade 2, being the most frequent adverse event. There were four grade 3 or greater adverse events, occurring once each in four patients: anemia, hyponatremia, aspiration pneumonia, and neutropenia, the last of which resolved after a 3-day interruption and resumption of the drug at half the original dose (80 mg) daily.

Dr. Soria noted that while the data look very promising, the real test of AZD9291 will come with longer follow-up showing whether the response is durable.

The study was funded by AstraZeneca. Dr. Lee is an advisory board member of Bristol-Myers Squibb, Korea, Boehringer Ingelheim Korea, and Pfizer Korea, and has received lecture fees from AstraZeneca, Ely Lilly, Merck Sharp & Dohme, and Roche. Dr. Soria reported having no disclosures.