Advertisement

Everything Is Connected

      Three decades ago, near the end of a 30-hour, minimal-sleep, San Francisco-Cleveland round trip, I fell into a deep sleep at 36,000 feet, awakened refreshed somewhere over Wyoming, and was rested enough for once to read Nature with care. To my great good fortune, it was the issue that reported the identification of RB, the gene whose mutations cause the eye tumor retinoblastoma (
      • Friend S.H.
      • Bernards R.
      • Rogelj S.
      • et al.
      A human DNA segment with properties of the gene that predisposes to retinoblastoma and osteosarcoma.
      ). The approach that led to this discovery was two pronged: (i) searching tumor cell DNA for areas of recurrent loss of heterozygosity (ie, loss of a portion of one of the two copies of a chromosome) that might include a putative tumor suppressor gene and (ii) doing family linkage studies on DNA gathered from kindreds with retinoblastomas. Because I have had a long-term interest generally in heritable disorders of the skin and identification of their molecular underpinnings and specifically in the basal cell nevus (Gorlin) syndrome (BCNS) (
      • Aurbach G.D.
      • Marcus R.
      • Winickoff R.N.
      • et al.
      Urinary excretion of 3′,5′-AMP in syndromes considered refractory to parathyroid hormone.
      ), the analogy between the clinical findings in retinoblastoma and basal cell carcinomas struck me as obvious—both came in two types: (i) the more common with a single sporadic tumor at a later age of onset and (ii) a rare variant, inherited as an autosomal dominant, often with multiple tumors and with an earlier age at onset. I thought that if Friend and colleagues could use their approach to identify the RB gene whose mutations underlie these tumors, we could use the same approach to identify the BCNS “gene,” and that might tell us something about the molecular aberration that causes the far more common sporadic basal cell carcinomas (BCCs).
      I immediately embarked on a project to gather many, many blood samples from families with BCNS so that we could do family linkage analysis and lots of BCCs so that we could do many, many Southern blots to identify the locus of the putative tumor suppressor gene that is aberrant in BCNS patients. We then discovered that groups on several continents had already embarked on this quest. In fact, after we had struggled for several years, the group led by Alan Bale got there first and, using the same approach, localized the gene for which we had been searching to chromosome 9q (
      • Gailani M.R.
      • Bale S.J.
      • Leffell D.J.
      • et al.
      Developmental defects in Gorlin syndrome related to a putative tumor suppressor gene on chromosome 9.
      ). The next step was to find the actual gene, and we and the rest of the community focused our efforts on an increasingly small area that “must” contain the gene. We were, as Lyndon Johnson put it, knee deep in the Big Muddy—lost in a forest with impenetrable fog that got thicker by the moment when, fortunately, an aha moment came. Matt Scott at Stanford called in October 1995 to tell me he thought he “had” the gene for which we had been searching for nearly a decade. Matt is a superb developmental biologist with a long record of contributions to elucidating the hedgehog (HH) signaling pathway; he is now president of the Carnegie Institution for Science. His lab had cloned the human homolog of the Drosophila gene encoding ptch, the primary inhibitor of that pathway, and David Cox and Richard Myers, human geneticists at Stanford, localized it to 9q; sure enough, there was a human disease, BCNS, at that site. Fortunately, he called us. The collaboration cleared the fog in which we were searching for the gene, and eight months later we published our findings of inactivating mutations in PTCH1 in BCNS patients and in DNA from sporadic and inherited BCCs (
      • Johnson R.L.
      • Rothman A.L.
      • Xie J.
      • et al.
      Human homolog of patched, a candidate gene for the basal cell nevus syndrome.
      ). Oh yes, on the very same day in June 1996, a multicontinent consortium, again led by Alan Bale, published their identification of the same gene, and they were smart enough to get there without needing a lucky phone call (
      • Hahn H.
      • Wicking C.
      • Zaphiropoulous P.G.
      • et al.
      Mutations of the human homolog of Drosophila patched in the nevoid basal cell carcinoma syndrome.
      ).
      The next lucky phone call was from Fred de Sauvage, at that time a young researcher and now vice president of Research Molecular Oncology at Genentech. His lab had cloned the human SMO gene, which encodes the protein that functions as the next step in the pathway, and together we found activating mutations in this gene (
      • Xie J.
      • Murone M.
      • Luoh S.M.
      • et al.
      Activating Smoothened mutations in sporadic basal-cell carcinoma.
      ), thus nailing the concept that it is elevated HH signaling that underlies all BCCs. Matt Scott’s lab made a Gorlin mouse (ie, Ptch1+/-) to study more fully the role of hedgehog (HH) signaling in development. He graciously gave it to us for studies of murine BCC carcinogenesis; we now have produced untold numbers of these furry beasts, all descended from one genetically engineered mouse, and they have been the linchpins of all our lab’s work for the past 1½ decades.
      But—except for enabling prenatal diagnosis—what was the utility of that discovery? At least theoretically, it opened the door to making a drug that might shut down the aberrant signaling pathway and thereby at least treat or, dare we hope, even cure BCCs. Scientifically, the way seemed clear, and Fred de Sauvage championed this project at Genentech. Alas, the company’s leadership, like that at most pharmaceutical companies, decided that BCCs were so well treated surgically that it was not worth the investment—better to pursue medical needs that seemed more unmet. Fortunately, data published in 2003 and 2004 suggested that enhanced HH signaling might be responsible for as many as 25% of all visceral cancers (eg,
      • Watkins D.N.
      • Berman D.M.
      • Burkholder S.G.
      • et al.
      Hedgehog signalling within airway epithelial progenitors and in small-cell lung cancer.
      ;
      • Berman D.M.
      • Karhadkar S.S.
      • Maitra A.
      • et al.
      Widespread requirement for Hedgehog ligand stimulation in growth of digestive tract tumours.
      ), and after that Pharma became highly eager to develop an HH inhibitor. Genentech partnered and eventually absorbed the HH inhibitor program that had been undertaken with the leadership of Lee Rubin (now a Harvard professor of stem cell biology) at Curis using a high-throughput screen of small molecules.
      The first such to be tested in vivo was Curis 61414, which has strong anti-BCC efficacy when applied to mouse skin but failed completely when applied to human skin (
      • Tang T.
      • Tang J.Y.
      • Li D.
      • et al.
      Targeting superficial or nodular basal cell carcinoma with topically formulated small molecule inhibitor of smoothened.
      ). Although the reason for this disparity is not completely understood, at least in part it was because of the perhaps order-of-magnitude greater barrier function in human compared with mouse skin and the drug's intrinsically stronger inhibitory effect on the murine than on the human molecular target. This failure encouraged Genentech to focus its efforts on developing an oral drug. The fruit of these efforts was vismodegib (Erivedge), which in 2012 became the first HH inhibitor to be approved by the US Food and Drug Administration (FDA) for marketing for treatment of “advanced BCCs,” an evolutionary term currently used to describe those BCCs for which surgery is at best a poor option. The good news is that it has remarkable antihuman BCC efficacy—approximately half of the very rare metastatic BCCs and the unusual but less rare locally advanced BCCs respond significantly (
      • Sekulic A.
      • Migden M.R.
      • Oro A.E.
      • et al.
      Efficacy and safety of vismodegib in advanced basal-cell carcinoma.
      ;
      • Basset-Seguin N.
      • Sharpe H.J.
      • de Sauvage F.J.
      Efficacy of Hedgehog pathway inhibitors in basal cell carcinoma.
      ). Genentech entrusted us with funds and medication during their phase II development of the drug, unusually early for such entrustment, enabling us to prosecute an investigator-initiated, double-blind, placebo-controlled trial of vismodegib’s anti-BCC efficacy in patients with Gorlin syndrome. We found that Gorlin BCCs essentially all melt away, eventually completely, and no new BCCs develop while patients remain on the drug. While participating in our trials, no patient has required excision of a BCC (
      • Tang J.Y.
      • Mackay-Wiggan J.M.
      • Aszterbaum M.
      • et al.
      Inhibiting the hedgehog pathway in patients with the basal-cell nevus syndrome.
      ). Unfortunately, vismodegib has systemic adverse effects that are class specific (hair loss that is likely due to the requirement for hedgehog signaling in anagen, taste loss that is likely due to the requirement for hedgehog signaling for taste bud development, and muscle cramps due to uncertain mechanisms).
      These adverse effects have been seen with other clinically studied hedgehog inhibitors, including sonidegib (Odomzo), the Novartis HH inhibitor that was approved by the FDA in July 2015 for sale in the US. Because of these side effects, many patients stop taking the drug, and when Gorlin patients stop taking the drug, BCCs that are histologically and clinically cured recur in the same site covering the same skin surface. Fortunately, when the drug is restarted, the tumors remain sensitive, and we have not seen any nonadvanced BCC develop resistance to vismodegib in nearly six years of study. Unfortunately, all publicly reported clinical trials of the efficacy of vismodegib and of the other small-molecule HH inhibitors versus other cancers (eg, colorectal, pancreatic, and ovarian cancers) have had disappointing results—BCCs and the subset of medulloblastomas that are hedgehog driven may be the only human tumors for which the current class of HH inhibitors has therapeutic efficacy.
      I draw several lessons from my participation in the identification of the molecular target, testing a topical preparation on mouse BCCs, and testing the oral drug on humans:
      • 1.
        Hillary’s speech writers were right—it takes a village, and in this case several villages—some with village elders such as J.B. Howell and other dermatologists who began describing patients with what now is termed BCNS in the 1950s; Robert Gorlin, who (like Columbus) was the “last” to describe the syndrome and therefore has the honor of having his name attached; Eric Wieschaus and Christiane Nüsslein-Volhard, the Nobel laureates who described the hedgehog signaling pathway in 1980; and hunter-gatherers such as the near dozen labs that worked collaboratively and competitively in the 1980s and 1990s to identify the gene whose mutations cause BCNS. Success also requires one or more of the near completely separate villages (such as Genentech or Novartis) that have the resources to fund the highly specialized armies of medicinal chemists, regulatory personnel, clinical trialists, and many others needed to move from concept and opportunity to supplying an effective drug to the shelves of pharmacies around the world.
      • 2.
        It takes a lot of money to feed and maintain even the last of these villages. The National Institutes of Health traditionally has supported the first villages. But these villages need enormous amounts of funds to develop new drugs—the estimates continue to climb from the hundreds of millions to more than a billion—and hence pharmaceutical companies need to invest in drugs that have at least a chance of returning large amounts of money.
      • 3.
        Given that the former villages have produced and no doubt will continue to produce more targets whose “hitting” could significantly ameliorate various diseases of the skin, how can the interest of the latter villages be piqued sufficiently to unleash the large amounts of capital needed to bring such a drug to market? That’s really the big question. Two types of answers come to mind.
        • i.
          In the case of melanoma, and increasingly for psoriasis, the market opportunity is perceived as being large enough to develop drugs whose first indication is dermatologic, although many would consider melanoma once it has left the skin an oncologic rather than a dermatologic disease. Can we find more dermatologic markets (numbers of patients × possible profit per patient) that are large enough that their targeting would justify the large amounts of capital that would have to be spent to develop new drugs specifically for those targets? In the case of the healing of chronic wounds and of the safe amelioration of eczema, the answer clearly would be affirmative. Perhaps this also might be the case for several orphan disease drugs with primarily skin manifestations because orphan diseases, at least for the time being, can command high prices per patient treated, and their path to approval may be less costly. But for the great majority of conditions to be found in any dermatologic text, the answer clearly is negative.
        • ii.
          In the case of hedgehog inhibitors for BCCs, it seems highly unlikely that the monies ever would have been spent were it to have been known that BCCs, and only a small subset of them (“advanced” or metastatic), would be the only market for the new drugs. The BCC drug came along on the coattails of anticipated indications for pancreatic, colorectal, and other cancers. This is a time-honored path for new dermatologic drugs (ie, using a drug developed for some other indication for a skin problem)—consider aminopterin and methotrexate decades ago and etanercept and other tumor necrosis factor inhibitors more recently for psoriasis, brimonidine for a red face, and steroids for almost everything. Could we repurpose newer drugs (or even old standbys) developed and utilized for other ailments for skin problems? Presumably to do so we would need to think more seriously about identifying such drug-skin targets and to institute a more structured approach to the problem. One example comes from the hedgehog inhibitor field. An in vitro targeted assay by Philip Beachy, then at Hopkins, of a large number of FDA-approved drugs identified both in vitro and in vivo (at least in our Ptch1+/- mice) antihedgehog/anti-BCC activity of itraconazole unrelated to its antifungal activity (
          • Kim J.
          • Tang J.Y.
          • Gong R.
          • et al.
          Itraconazole, a commonly used antifungal that inhibits Hedgehog pathway activity and cancer growth.
          ,
          • Kim J.
          • Aftab B.T.
          • Tang J.Y.
          • et al.
          Itraconazole and arsenic trioxide inhibit Hedgehog pathway activation and tumor growth associated with acquired resistance to smoothened antagonists.
          ). Investigation of its anti-HH efficacy in humans, including its anti-prostate cancer efficacy, is under way. Its potency is considerably less than that of vismodegib and other “professional” hedgehog inhibitors but if we did not have the latter, perhaps it could have served as an at least partially effective anti-BCC treatment. The regulatory and financial hurdles blocking repurposing of an old drug are far lower than are those blocking development of a new drug. Are there more such in vitro screens of already approved drugs that might unearth yet other, even more useful surprises?
      • 4.
        I have been extraordinarily fortunate to have seen work that our lab has done actually lead to something useful for patients. How many people get to work on the identification of the molecular basis of a disease and then lead the clinical trial of the first drug to replace the function of the defective gene? What underlies that good fortune, beyond a very healthy dose of good luck? I could have skipped the Cleveland trip and not have read about RB; Matt could have called someone else; we could have worked on a disease for which the identification of the mutant gene did not lead so clearly to a drug. Indeed, our own findings of keratin gene mutations in epidermolysis bullosa simplex have not enabled a therapy. Part of it clearly depends on my having been surrounded from the start by very good individuals. Because of the example set by my family, I grew up expecting to become a clinical scholar in an era in which the phrase “balanced life” had not yet been coined and never felt that work was drudgery. And my lab at the San Francisco General Hospital for 35 years was adjacent to that of Y.W. Kan, a pioneer in applying molecular biology to clinical problems; he was kind enough to teach us the fundamentals of this then-arcane field. More broadly, being immersed even half-time in a vibrant research university gave me exposure to new ideas and the sense that if these guys could employ new approaches, maybe so could I. Part of it depended on having the freedom to travel. For many years I was not only a faithful attendee at the meetings of the Society for Investigative Dermatology but also the sole card-carrying dermatologist at the annual meeting of the American Society for Human Genetics. Because of this, in 1980 I was, along with Alain Hovnanian, one of the two persons on the planet who knew something about two particular fields of knowledge—family linkage analysis and Mendelian diseases of the skin. And part of it was because I had both a clinical practice and a lab that allowed me to approach whatever problem I wanted so that I could fantasize about how wonderful it would be if someday I could go to some far off country and see on the druggist’s shelf a drug whose development I had touched, even peripherally. It happened, and the patients are even happier than I am. Now if only we can convert remission into cure!

      REFERENCES

        • Aurbach G.D.
        • Marcus R.
        • Winickoff R.N.
        • et al.
        Urinary excretion of 3′,5′-AMP in syndromes considered refractory to parathyroid hormone.
        Metabolism. 1970; 19: 799-808
        • Basset-Seguin N.
        • Sharpe H.J.
        • de Sauvage F.J.
        Efficacy of Hedgehog pathway inhibitors in basal cell carcinoma.
        Mol Cancer Ther. 2015; 14: 633-641
        • Berman D.M.
        • Karhadkar S.S.
        • Maitra A.
        • et al.
        Widespread requirement for Hedgehog ligand stimulation in growth of digestive tract tumours.
        Nature. 2003; 425: 846-851
        • Friend S.H.
        • Bernards R.
        • Rogelj S.
        • et al.
        A human DNA segment with properties of the gene that predisposes to retinoblastoma and osteosarcoma.
        Nature. 1986; 323: 643-646
        • Gailani M.R.
        • Bale S.J.
        • Leffell D.J.
        • et al.
        Developmental defects in Gorlin syndrome related to a putative tumor suppressor gene on chromosome 9.
        Cell. 1992; 69: 111-117
        • Hahn H.
        • Wicking C.
        • Zaphiropoulous P.G.
        • et al.
        Mutations of the human homolog of Drosophila patched in the nevoid basal cell carcinoma syndrome.
        Cell. 1996; 85: 841-851
        • Johnson R.L.
        • Rothman A.L.
        • Xie J.
        • et al.
        Human homolog of patched, a candidate gene for the basal cell nevus syndrome.
        Science. 1996; 272: 1668-1671
        • Kim J.
        • Tang J.Y.
        • Gong R.
        • et al.
        Itraconazole, a commonly used antifungal that inhibits Hedgehog pathway activity and cancer growth.
        Cancer Cell. 2010; 17: 388-399
        • Kim J.
        • Aftab B.T.
        • Tang J.Y.
        • et al.
        Itraconazole and arsenic trioxide inhibit Hedgehog pathway activation and tumor growth associated with acquired resistance to smoothened antagonists.
        Cancer Cell. 2013; 23: 23-34
        • Sekulic A.
        • Migden M.R.
        • Oro A.E.
        • et al.
        Efficacy and safety of vismodegib in advanced basal-cell carcinoma.
        N Engl J Med. 2012; 366: 2171-2179
        • Tang T.
        • Tang J.Y.
        • Li D.
        • et al.
        Targeting superficial or nodular basal cell carcinoma with topically formulated small molecule inhibitor of smoothened.
        Clin Cancer Res. 2011; 17: 3378-3387
        • Tang J.Y.
        • Mackay-Wiggan J.M.
        • Aszterbaum M.
        • et al.
        Inhibiting the hedgehog pathway in patients with the basal-cell nevus syndrome.
        N Engl J Med. 2012; 366: 2180-2188
        • Watkins D.N.
        • Berman D.M.
        • Burkholder S.G.
        • et al.
        Hedgehog signalling within airway epithelial progenitors and in small-cell lung cancer.
        Nature. 2003; 422: 313-317
        • Xie J.
        • Murone M.
        • Luoh S.M.
        • et al.
        Activating Smoothened mutations in sporadic basal-cell carcinoma.
        Nature. 1998; 391: 90-92