Lack of trastuzumab activity in nonsmall cell lung carcinoma with overexpression of erb-B2

39810: A Phase II trial of cancer and leukemia group B


  • The following institutions participated in this study: CALGB Statistical Center, Durham, NC (Stephen George, Ph.D.; supported by grant CA336) Christiana Care Health Services, Inc. Community Cooperative Oncology Program (CCOP), Wilmington, DE (Stephen Grubbs, M.D; supported by grant CA45418); Dartmouth Medical School–Norris Cotton Cancer Center, Lebanon, NH (Marc S. Ernstoff, M.D.; supported by grant CA04326); Duke University Medical Center, Durham, NC (Jeffrey Crawford, M.D.; supported by grant CA47577); Greenville CCOP, Greenville, SC (Jeffrey Giguere, M.D.); Long Island Jewish Medical Center, Lake Success, NY (Marc Citron, M.D.; supported by grant CA11028); Rhode Island Hospital, Providence, RI (William Sikov, M.D.; supported by grant CA08025); Southeast Cancer Control Consortium Inc. CCOP, Goldsboro, NC (James N. Atkins, M.D.; supported by grant CA45808); State University of New York Upstate Medical University, Syracuse, NY (Stephen L. Graziano, M.D.; supported by grant CA21060); and Washington University School of Medicine, St. Louis, MO (Nancy Bartlett, M.D.; supported by grant CA77440).

  • The content of this article is solely the responsibility of the authors and does not necessarily represent the official views of the National Cancer Institute.



The overexpression of HER-2 occurs in a minority of patients with nonsmall cell lung carcinoma. Trastuzumab, which is a monoclonal antibody to HER-2, is an effective treatment in some women with breast carcinomas that overexpress HER-2, as demonstrated by immunohistochemistry. The objective of this Phase II study was to determine whether trastuzumab would effect responses in patients with nonsmall cell lung carcinoma who had tumors that overexpressed HER-2.


Patients were required to have Stage IIIB or Stage IV nonsmall cell lung carcinoma and tumors with 2+ or 3+ expression of HER-2, as determined with immunohistochemistry, and they may have received up to 1 prior chemotherapy regimen. Trastuzumab at a dose of 4 mg/kg was given intravenously on Week 1; then, weekly doses of 2 mg/kg were given. Response revaluation was performed every 8 weeks.


Among 209 screened patients, 24 patients (11%) had tumors with 2+ or 3+ expression of HER-2. One patient achieved a partial response, and one patient experienced a treatment-related death due to pulmonary toxicity.


Single-agent trastuzumab did not exhibit significant clinical activity against nonsmall cell lung carcinoma when HER-2 expression levels were measured by immunohistochemistry. Cancer 2005. © 2005 American Cancer Society.

HER-2 (also called c-erb-B2 or neu) is a tyrosine kinase receptor that mediates proliferation and differentiation in normal epithelial cells. The overexpression of HER-2 contributes to tumor growth.1 In patients with breast carcinoma, the overexpression of HER-2 has been associated with a poorer prognosis and with response to therapy with an antibody to HER-2 (trastuzumab).2 In women with breast carcinoma, therapy with trastuzumab is important clinically both in the setting of metastatic disease and in the adjuvant setting.3–5 Trastuzumab enhances the antitumor activity of a number of types of chemotherapy in patients with breast carcinoma.5

Estimates of the frequency of overexpression of HER-2 in nonsmall cell lung carcinoma (NSCLC) ranged from 2% to 40% in early studies.1 In 1990, Kern et al. first demonstrated overexpression of HER-2 in ≈ 15% of NSCLCs.6 HER-2 overexpression in adenocarcinoma of the lung also predicted a worse survival despite curative treatment compared with similar patients who had adenocarcinoma that did not overexpress HER-2.6 A direct role of HER-2 in the proliferation of NSCLC was postulated when an antibody to HER-2 inhibited the growth of NSCLC cell lines that overexpressed HER-2.7 In a more recent trial, these observations were supported with the finding that amplification of HER-2 in NSCLC was a negative prognostic factor in patients who had undergone surgical resection.8 To determine the single-agent efficacy of trastuzumab in patients with NSCLC who had tumors that overexpressed HER-2, the Cancer and Leukemia Group B (CALGB) initiated a trial in February 2000 that included patients with Stage IV disease and patients with Stage IIIB disease whose extent of disease precluded an attempt at curative therapy with chemotherapy and/or radiation therapy. The objective of this study was to examine the efficacy of trastuzumab within 2 patients groups with tumors that overexpressed HER-2: Patients who had received no prior chemotherapy and patients who had been treated with a single prior chemotherapy regimen.

To assure quality control in the assessment of HER-2, all tissue specimens were sent to a central core laboratory (LabCorp, Research Triangle Park, NC). Recognizing that ≈ 8–9 negative specimens would need to be submitted for each patient who had NSCLC that exhibited 2–3+ expression, it was estimated that the sample size goal of ≈ 39 patients with tumors that overexpressed HER-2 within each of 2 patient groups required screening a total of 560 patients for overexpression (260 per patient group). After 29 months of patient accrual, 209 patients had been screened, and only 24 patients (11.5%) had been registered formally as eligible to receive trastuzumab. The study was closed at that time. We report here on the efficacy of trastuzumab as single-agent therapy in the 22 eligible patients with NSCLC who had tumors that overexpressed HER-2.


In this multicenter, nonrandomized trial, patients were required to have advanced (AJCC Stage IIIB or Stage IV) NSCLC with measurable disease. Patients also were required 1) to provide informed consent; 2) to be age ≥ 18 years; 3) to have a performance score of 0–2; 4) to have no underlying immunologic disease; 5) not to be pregnant or nursing; 6) to have 4 weeks since any prior chemotherapy or radiation therapy; 7) to be at least 4 weeks since any major surgery; and 8) to have a granulocyte count > 1000/μL, a platelet count > 60,0000/μL, creatinine < 1.5 × the upper limit of normal (ULN), bilirubin < 1.5 × ULN, and a left ventricular ejection fraction (LVEF) as determined by echo or multiple-gated acquisition scan of > 45%. Patients were excluded for 1) the presence of central nervous system metastases; 2) the presence of any prior anthracycline therapy; and 3) greater than one prior chemotherapy for lung carcinoma. Patients who had tumors that expressed HER-2 at the 2–3+ level and fit the above inclusion and exclusion criteria were registered on the study. Like patients with breast carcinoma, patients who had 0–1+ expression of HER-2 by immunohistochemistry were considered negative.

Pretreatment evaluation included assessment of the tumor for HER-2 expression; a history and physical examination; record of tumor size and performance status; complete blood count and differential; serum chemistries, including aspartate aminotransferase, alkaline phosphatase, bilirubin, albumin lactate dehydrogenase, serum glucose and electrolytes, serum blood urea nitrogen and creatinine, and a complete blood count with differential. A chest X-ray and computed tomography (CT) scan of the chest to include the liver and adrenals were required prior to study. Other staging studies, such as a bone scans or brain magnetic resonance images, were obtained as indicated clinically. Blood work and toxicity assessments were repeated every 4 weeks during therapy, and tumor response was evaluated every 8 weeks.

The study was performed according to the Declaration of Helsinki and was approved by the executive committee of CALGB, the National Cancer Institute-Cancer Therapy Evaluation Program review committee, and the individual institutional review boards of the participating institutions. Written informed consent was obtained from all participants. Centralized data review was performed to ensure completeness and quality prior to statistical analysis at the CALGB statistical center at Duke University. The CALGB data audit committee reviewed much of the primary data and X-rays during the course of their institutional audits. For patients in whom there was a question about response, the relevant CT scans were reviewed at multiple centers.

Protocol Treatment

Patients were treated with trastuzumab (Herceptin®; Genentech Inc., South San Francisco, CA). The first infusion was 4 mg/kg given over 90 minutes; then, weekly infusions of 2 mg/kg were given over 30 minutes. To manage acute infusion-related toxicities, the rate of infusion could be reduced. Pretreatment with acetaminophen and/or diphenhydramine was suggested. Chills related to infusion could be treated with meperidine. Dose modifications were based on any observed toxicity. Four weekly treatments were considered one cycle of therapy. Every 8 weeks (2 cycles), an evaluation for response was indicated. All patients had to have a follow-up assessment of LVEF every 3 months during therapy. Therapy was to be continued until patients developed disease progression or stopped if unacceptable toxicity was observed.

Criteria of Response

For response criteria, we used the National Cancer Institute-designated Response Evaluation Criteria in Solid Tumors.8, 9 To be assigned a status of partial response or complete response, change in tumor measurements had to be confirmed by repeat studies no less than 4 weeks after the initial criteria for response first were met. Furthermore, in patients with stable disease, follow-up measurements had to meet the stable disease criteria at least once after study entry at an interval of 8 weeks. The duration of response was calculated from the time measurement criteria for response first were met to the time of disease progression.

Statistical Methods

The primary objective was to ascertain the activity of trastuzumab among 2 patient groups with NSCLC who had tumors that exhibited 2+ or 3+ expression: patients with no prior treatment and patients with a single prior chemotherapy treatment. Within each patient subgroup, the objective was to differentiate between a response rate ≤ 5% and a response rate > 20%. A 2-stage design was used so that, if 0–1 responses were observed in the first 22 patients, then the study would be terminated early. The Type I and II error rates were < 0.10. The product-limit estimator developed by Kaplan and Meier was used to illustrate survival and failure-free survival. Survival was defined as the time between initiation of protocol treatment and death, and failure-free survival was defined as the time between initiation of protocol treatment and first failure (death, disease progression, or recurrence). Toxicity was summarized by type and worst episode per patient.


Between February 15, 2000 and July 19, 2002, 209 patients were screened. Twenty-four patients with tumors that exhibited 2+ or 3+ overexpression (11%) were registered for the study. Two patients were not evaluable: one patient who never received any treatment and one patient who had more than a single prior treatment regimen. The latter patient was considered ineligible. All 23 patients who received any therapy were included in the analysis of toxicity, but the 22 eligible patients were included in the analyses of response, survival, and failure-free survival.

Patient Characteristics

Table 1 summarizes the characteristics of the 22 eligible patients. There were 64% males, and the median age was 65 years. The performance score was 0–1 in 91% of patients. However, weight loss > 5% was noted in 27% of patients at the time of study entry. Of the patients who were treated, 1 patient had 3+ expression of HER-2 by immunohistochemistry, and all remaining patients had 2+ expression. The median number of weekly doses of trastuzumab received was 8, with a range of 1–27 weeks. Only four patients had less than one cycle of therapy.

Table 1. Pretreatment Characteristics
CharacteristicNo. of patients (%)
No prior treatmentPrior treatmentAll patients
 Male7 (78)7 (54)14 (64)
 Female2 (22)6 (46)8 (36)
 Median (yrs)735965
 Range (yrs)  45–81
 40–49 yrs0 (0)3 (23)3 (14)
 50–59 yrs0 (0)4 (31)4 (18)
 60–69 yrs1 (11)4 (31)5 (23)
 70–79 yrs7 (78)2 (15)9 (41)
 ≥ 80 yrs1 (11)0 (0)1 (5)
Performance status   
 02 (22)7 (54)9 (41)
 15 (56)6 (46)11 (50)
 22 (22)0 (0)2 (9)
Weight loss   
 < 5%4 (44)9 (69)13 (59)
 > 5%3 (33)3 (23)6 (27)
 Unknown2 (22)1 (8)3 (14)


One patient suffered a likely treatment-related death. This patient had received prior chemotherapy and radiation therapy and reportedly had chest X-ray findings that were consistent with pneumonitis prior to trastuzumab therapy. However, the patient developed rapidly progressive infiltrates and respiratory failure consistent with acute drug-induced pneumonitis. After this incident was reported, accrual to the study was suspended temporarily, and further accrual was limited to patients with no prior chest irradiation. Two other patients (9%) reportedly had Grade 3 hypoxia, and 1 other patient reportedly had Grade 1 pneumonitis.

The patient who developed fatal pulmonary toxicity had received prior radiation of up to 4500 centigrays to the tumor with concurrent weekly taxol and carboplatin. The last radiation dose was given 3 months prior to the administration of the first dose of trastuzumab. This patient had tumor in the contralateral lung, the liver, the adrenals, and the pancreas at the time trastuzumab therapy was started. The patient had a normal LVEF of 55% at the initial administration of trastuzumab. The patient received 1 dose of trastuzumab on March 11, 2000; and, by March 14, 2000, a CT scan showed diffuse, ground-glass opacities in both lung fields. The patient died of respiratory failure on March 20, 2004. The patient was on nine other medications and vitamins, but none were associated with pulmonary toxicity.

Other serious toxicities observed included 1 patient (4%) with a decrease in LVEF from 52% to 45%. Two patients (9%) suffered from fatigue. One patient each was described with Grade 3 rigors, constipation, hyponatremia, muscle weakness, and hallucinations.

Response to Treatment

One patient who had received prior chemotherapy did achieve an objective partial response. The response rate was 5%, with a 95% confidence interval of 1–23%. Six patients (27%) had a decrease in the size of measurable lesions, but those decreases were insufficient to meet the criteria for a partial response. One other patient had stable disease. The duration of response to therapy was a median of 2.6 months, as shown in Table 2. The reasons for stopping therapy were progressive disease in 14 patients (64%), death due to toxicity in 1 patient (5%), toxicity in 2 patients (9%), withdrawal by 1 patient (5%), and a change of therapy in 1 patient (5%). In three patients, the off-treatment reason was not related to toxicity or disease progression.

Table 2. Survival Based on Prior Therapy
Survival characteristicsNo. of patientsNo. of eventsMedian survival (mos)95% CI (mos)
  1. 95% CI: 95% confidence interval.

Overall survival22205.34.2–12.6
 With no prior treatment976.3 
 With prior treatment13135.3 
Failure-free survival22222.62.2–4.2
 With no prior treatment992.8 
 With prior treatment13132.2 

The one patient who had an objective response had 2+ expression of HER-2 by immunohistochemistry. An analysis of HER-2 by fluorescent in situ hybridization was not obtained. This patient had an adenocarcinoma of the lung. At the start of trastuzumab therapy, the disease was spread within the lung, and there were bilateral pleural effusions and bone metastases. The pretreatment performance status was 0, and the patient had not experienced any weight loss. The patient had received prior chemotherapy with paclitaxel and carboplatin. The patient received weekly therapy from February 28, 2001 to August 15, 2001, which comprised 6 cycles of therapy plus 1 other weekly dose. On August 21, 2002, it was found that the patient had progressive disease. The patient died 9 months after the start of trastuzumab therapy. Overall survival and failure-free survival are presented in Figures 1 and 2, respectively.

Figure 1.

Overall failure-free survival in the Cancer and Leukemia Group B (CALGB) 39810 trial.

Figure 2.

Overall survival in the Cancer and Leukemia Group B (CALGB) 39810 trial.


The failure to observe > 1 response in the first 22 evaluable patients and the serious toxicity observed led to the early closure of this CALGB trial. In patients with metastatic breast carcinoma, immunohistochemical demonstration of 2+ or 3+ staining for HER-2 predicts an ≈ 25% response rate to therapy with trastuzumab.2 Although it was reasonable to conjecture that trastuzumab may be effective therapy in the relatively small cohort of patients with NSCLC who have tumors that overexpress HER-2, we did not observe more than a minimal response in the current trial. Although other investigators have evaluated trastuzumab in combination with other therapy, we are unaware of other studies examining the single-agent activity of trastuzumab in NSCLC. The results of the current study suggest very strongly that the utility of single-agent trastuzumab is so minimal in lung carcinoma patients with 2+ expression of HER-2 that further studies in this population are not warranted. There was only 1 patient in our study who had 3+ expression. Although there may be more utility for trastuzumab in patients with 3+ expression of HER-2, such patients represent < 2% of patients with NSCLC.

Early studies of trastuzumab in combination with chemotherapy for NSCLC have not established that trastuzumab adds to therapy despite low single-agent activity.11, 12 There is a suggestion that, in the few patients with 3+ HER-2 expression by immunohistochemistry, survival may be slightly better using trastuzumab as part of therapy.12, 13 Because the number of patients with lung carcinoma who have 3+ HER-2 expression is so low, it would require screening thousands of patients to try to carry out a study for just this subgroup of patients. There certainly are other more common molecular targets in patients with NSCLC (mutations in the epidermal growth factor receptor [EGFR] and in K-ras) that are more relevant clinically for study at this time.

Increased activity of the EGFR superfamily occurs with a number of different molecular events.14 The most common defect is overexpression of the EGFR ligands. Another common defect is a mutation that leads to persistently active receptor. With mutation, dimers of HER-2 and EGFR also occur. It is not certain that the molecular changes in NSCLC are the same as those seen in breast carcinoma. The recent work of Lynch et al.15 describing the type of EGFR1 mutations that govern responsiveness to gefitinib will allow the majority of patients who will respond to gefitinib to be identified prior to therapy, sparing other patients the cost and time of using ineffective therapy. Prior to further work on trastuzumab in lung carcinoma, the molecular changes associated with response will need to be identified. If only some of the < 2% of patients with lung carcinoma who have 3+ HER-2 expression benefit from trastuzumab therapy, and if the magnitude of that benefit may be measured in a few months' improvement in survival without cure, then screening the ≥ 98% patients who will not respond is difficult to justify.

The toxicity observed in this trial included one patient with a drop in LVEF and one patient who developed fatal pulmonary toxicity. Although cardiac toxicity is a well described complication of trastuzumab,16 severe pulmonary toxicity has not been described well in the literature. The package insert for trastuzumab does contain a warning that pulmonary toxicity occurs and can be severe. Documentation of lung toxicity is difficult to find in the literature. Radzikowska et al. described a patient with reversible lung toxicity when the trastuzumab was held.17 It is unclear whether prior lung radiation or underlying radiation pneumonitis predisposes to lung toxicity or whether this is an idiosyncratic event.

Trastuzumab cannot be recommended at the standard dose and schedule used for patients with breast carcinoma in patients with NSCLC. Further definition of the molecular changes in those rare patients with HER-2 expression ≥ 3+ may predict response, but the clinical infrequency of such patients makes such studies impractical. Studies of trastuzumab in combination with chemotherapy for patients with HER-2 expression levels of 0–2+ do not appear justified.