[Skip to Content]
Access to paid content on this site is currently suspended due to excessive activity being detected from your IP address Please contact the publisher to request reinstatement.
[Skip to Content Landing]
Views 3,103
Citations 0
September 10, 2019

Biomarker-Driven Drug Discovery in Cancer—Trastuzumab Development: 2019 Lasker-DeBakey Clinical Medical Research Award

Author Affiliations
  • 1Biooncology Consultants, San Diego, California
JAMA. Published online September 10, 2019. doi:10.1001/jama.2019.13963

The 2019 Lasker-DeBakey Clinical Medical Research Award has been presented to H. Michael Shepard, Dennis J. Slamon, and Axel Ullrich for the invention of trastuzumab, the first monoclonal antibody that blocks a cancer-causing protein, and for its development as a life-saving therapy for women with breast cancer.

The trastuzumab discovery program started with the urgent unmet need to create cancer drugs that were safe for patients, but deadly for tumors. In the beginning, this work was not related to breast cancer. It had to do with the observation that drugs, such as carmustine (bis-chloroethylnitrosourea, BCNU, BiCNU), primarily used for glioblastoma, exerted their effects by dissolving cells. The premise of this work was that this sort of drug should not be the first treatment of choice for any disease, including cancer. Thus, in 1985, the initial approach was to find a frequent change, common among tumor cells, that would be essential to tumor progression, and thus would offer the possibility of being a cancer-specific therapy. The “magic bullet” idea was not novel, but aside from the possible exception of early-stage surgery, very few cancer treatments appeared to have this property. Examples of success include methotrexate, used to cure choriocarcinoma, and cisplatin, when it was used to cure many cases of testicular cancer. However, not all patients responded to these treatments and there was no way to predict which patients would respond. Thus, identifying biomarkers that would predict sensitivity to a particular drug also became a goal.

Tumor Necrosis Factor
Failure and a Way Forward

The initial efforts focused on a molecule called tumor necrosis factor (TNF or TNF-α), which had first been described by Lloyd Old and colleagues as a tumor toxic factor in the serum of endotoxin-treated mice. Interest in this molecule first arose in experiments by Old et al in which they either treated mice with endotoxin (which induced TNF) or took serum from endotoxin-treated mice and showed first that it was toxic to mouse L cells in vitro, and, amazingly, caused hemorrhagic necrosis and then regression of transplanted MethA sarcoma in mice.

Subsequently, David Goeddel and colleagues at Genentech were able to purify enough of this factor to obtain amino acid sequence, and then clone cDNA encoding human TNF. The TNF purified by Goeddel et al demonstrated the same properties as Old had described, but after being readied for clinical trials, the recombinant human TNF was very toxic when administered to patients. This halted its clinical development. Furthermore, to everyone’s surprise, when TNF was tested by Sugarman et al1 on a wide variety of tumor cell lines in vitro, it failed to display cytotoxic effects on more than two-thirds of the cell lines that were tested. With publication of the article by Sugarman et al describing the (lack of) effects of TNF on most tumor cell lines, Old wrote a flattering editorial about the findings, although the results did not support continued clinical testing of TNF.

The widespread resistance of tumor cells to TNF was not good news for its use as an anticancer agent in clinical trials, but suggested that TNF resistance may be an underlying property shared by many established tumors.

Macrophages and ERBB2

A key question was whether TNF was potentially a part of macrophage surveillance for incipient tumor cells. The answer to this question came as a result of a fortuitous collaboration with Hans Schreiber, whose earlier work had led to the suspicion that macrophages are a key factor in antitumor surveillance, that they might do it by producing TNF, and that tumor cells had to escape TNF/macrophages to achieve malignant progression. The collaboration showed that TNF was an effector molecule of macrophages, and that incipient tumor cells had to traverse this hurdle to become more aggressive.2 This study served to provide a platform for the hypothesis that resistance to TNF could be an early step in tumor progression. Also, because all (or nearly all) of the tumor cell lines in our early experiments were derived from biopsies of aggressive tumors, this hypothesis could explain why most human tumor cell lines from the American Type Culture Collection, which had already confronted the macrophage in vivo and survived, were already resistant to TNF-mediated cytotoxicity.

This led to thinking about tumor cell properties and actions that were already known to enable escape from normal growth controls. One mechanism that attracted attention was the autocrine growth hypothesis of Sporn and Todaro, which proposed that autocrine and paracrine production of growth factors from tumor cells could sustain their growth, making them independent of host-supplied factors. Sugarman and Gail Phillips proceeded to choose several different TNF-sensitive cell lines that were on hand and co-incubated them with growth factors and TNF. The results showed up to 95% inhibition of the cytotoxic effects of TNF with epidermal growth factor and several other growth factors.3 These important experiments helped define activated receptor tyrosine kinases as possible targets in the quest for a tumor-selective cancer therapeutic.

The next step came from interaction with Robert Hudziak, a postdoctoral fellow in Axel Ullrich’s laboratory at Genentech, who was very excited about the idea that TNF resistance could be a step in the evolution of tumor progression. Hudziak was working with a receptor tyrosine kinase that had been cloned a year before in the Ullrich laboratory, called ERBB2 (formerly HER2). Hudziak had generated cell lines that expressed increasing amounts of ERBB2, and, together with Bryan Fendly, had also generated a panel of antibodies to the extracellular domain of ERBB2 that could be used to screen for potential receptor antagonists. The results of collaboration to test these cell lines for their sensitivity/resistance to TNF and later studies showed that ERBB2 overexpression induced resistance to TNF.4 Furthermore, some of the monoclonal antibodies targeting the ERBB2 extracellular domain, especially one called 4D5, selectively inhibited tumor cell growth in vitro and in vivo and induced TNF sensitivity in ERBB2-overexpressing cell lines.5 Importantly, 4D5 would only do these things if the receptor was highly expressed in the range of 30-fold higher than in a normal fibroblast.6 This latter finding had huge implications for future therapy, including the concept of preselecting patients with overexpression of ERBB2 to enrich for those most likely to respond to treatment. The legacy of developing trastuzumab is not only that it was the first monoclonal antibody effective against solid tumors or that it was the first tyrosine kinase inhibitor to be approved, but also because it helped pioneer the era of biomarker-driven drug discovery and development.

These data—combined with work from Stuart Aaronson’s laboratory, which showed overexpression of ERBB2 in breast cancer, and the results from other studies, including the collaboration between Slamon and colleagues7 that showed overexpression of ERBB2 predicts shorter survival in breast and ovarian cancer—were very exciting. The results indicated that at least part of the extraordinary aggressiveness of ERBB2-overexpressing cancer was likely due to ERBB2 inducing resistance of cancers to host immune surveillance (TNF/macrophages). It was not just the growth-promoting function of the receptor tyrosine kinase, or a result of transformation, because in contrast to ERBB2-induced transformation, ras-transformed cells had unchanged sensitivity to TNF.

Onward to Trastuzumab

Having worked out that inhibition of function could interfere with multiple tumorigenic functions of ERBB2, and that its tumorigenic effect correlated with its overexpression in cancer, it was necessary to determine how to create a useful therapeutic, which was approached in 2 ways. First an active specific immunotherapy approach was taken. This provided encouraging early data, but the system was very complex. This led us to pursue the “simplicity” of monoclonal antibody therapy.8 For this purpose, a novel method for humanizing the 4D5 monoclonal antibody was invented by Carter et al,9 who coined the process “6-pack mutagenesis.” The antibody, then known as HuMAb4D5, was brought forward to clinical trials. The antibody has now been used to treat about 2 million women with ERBB2-overexpressing breast cancer, many of whom have received significant benefit. However, not all patients with ERBB2-overexpressing breast cancer respond, and acquired resistance does occur. For this reason, Phillips and colleagues at Genentech have created 2 follow-on anti-ERBB2 therapeutics. The first was pertuzumab, which binds to a different site than trastuzumab on the ERBB2 extracellular domain. Combinations of pertuzumab with trastuzumab have resulted in striking clinical results in some patients. More recently, a first-generation antibody-drug conjugate (ado-trastuzumab emtansine) has been introduced10 and is active in treatment of trastuzumab-resistant disease.

Lessons Learned From the Development of Trastuzumab

The work that led to this pioneering success in biomarker-driven cancer therapy required collaboration, collaboration, and collaboration: first between basic research scientists in academia, including Robert Weinberg, who first showed that receptor tyrosine kinases could be transforming; scientists at Genentech, including Artur Levinson and Ullrich, who were able to clone and characterize ERBB2; and then my group, also at Genentech, which provided a scientific rationale as to why ERBB2 could be a good target for cancer therapy. The research environment at Genentech, including ambition and a “take no prisoners” dedication to good science, was an essential contributor to the success of this project. Later, the involvement and early enthusiasm for developing trastuzumab by the clinical research community, most notably Dennis Slamon at the University of California, Los Angeles, and Larry Norton at Sloan Kettering, was essential. There has been much made of delays and potential abandonment of this program by Genentech at various times when hurdles appeared difficult to surmount. However, it was only 6 years from first patient enrolled to first approval, which was amazingly fast considering that this was the first fully humanized monoclonal antibody for the treatment of solid tumors, the first approved tyrosine kinase inhibitor, and the first successful biomarker-driven drug development program in cancer. It is now a model for nearly every new drug developed to treat cancer.

Back to top
Article Information

Corresponding Author: H. Michael Shepard, PhD, 13225 Bavarian Dr, San Diego, CA 92129 (HMS@betteroutcomes4cancer.com).

Published Online: September 10, 2019. doi:10.1001/jama.2019.13963

Conflict of Interest Disclosures: Dr Shepard reported receiving salary and stock from Genentech.

Sugarman  BJ, Aggarwal  BB, Hass  PE, Figari  IS, Palladino  MA  Jr, Shepard  HM.  Recombinant human tumor necrosis factor-alpha.  Science. 1985;230(4728):943-945.PubMedGoogle ScholarCrossref
Urban  JL, Shepard  HM, Rothstein  JL, Sugarman  BJ, Schreiber  H.  Tumor necrosis factor.  Proc Natl Acad Sci U S A. 1986;83(14):5233-5237.PubMedGoogle ScholarCrossref
Sugarman  BJ, Lewis  GD, Eessalu  TE, Aggarwal  BB, Shepard  HM.  Effects of growth factors on the antiproliferative activity of tumor necrosis factors.  Cancer Res. 1987;47(3):780-786.PubMedGoogle Scholar
Hudziak  RM, Lewis  GD, Shalaby  MR,  et al.  Amplified expression of the HER2/ERBB2 oncogene induces resistance to tumor necrosis factor alpha in NIH 3T3 cells.  Proc Natl Acad Sci U S A. 1988;85(14):5102-5106.PubMedGoogle ScholarCrossref
Hudziak  RM, Lewis  GD, Winget  M, Fendly  BM, Shepard  HM, Ullrich  A.  p185HER2 monoclonal antibody has antiproliferative effects in vitro and sensitizes human breast tumor cells to tumor necrosis factor.  Mol Cell Biol. 1989;9(3):1165-1172.PubMedGoogle ScholarCrossref
Lewis  GD, Figari  I, Fendly  B,  et al.  Differential responses of human tumor cell lines to anti-p185HER2 monoclonal antibodies.  Cancer Immunol Immunother. 1993;37(4):255-263.PubMedGoogle ScholarCrossref
Slamon  DJ, Clark  GM, Wong  SG, Levin  WJ, Ullrich  A, McGuire  WL.  Human breast cancer.  Science. 1987;235(4785):177-182.PubMedGoogle ScholarCrossref
Fendly  BM, Kotts  C, Vetterlein  D,  et al.  The extracellular domain of HER2/neu is a potential immunogen for active specific immunotherapy of breast cancer.  J Biol Response Mod. 1990;9(5):449-455.PubMedGoogle Scholar
Carter  P, Presta  L, Gorman  CM,  et al.  Humanization of an anti-p185HER2 antibody for human cancer therapy.  Proc Natl Acad Sci U S A. 1992;89(10):4285-4289.PubMedGoogle ScholarCrossref
Phillips  L, Li  G, Dugger  DL,  et al.  Targeting HER2-positive breast cancer with trastuzumab-DM1, an antibody-cytotoxic drug conjugate.  Cancer Res. 2008;68(22):9280-9290Google ScholarCrossref
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words