Development of Genome-Derived Tumor Type Prediction to Inform Clinical Cancer Care.

IMPORTANCE: Diagnosing the site of origin for cancer is a pillar of disease classification that has directed clinical care for more than a century. Even in an era of precision oncologic practice, in which treatment is increasingly informed by the presence or absence of mutant genes responsible for cancer growth and progression, tumor origin remains a critical factor in tumor biologic characteristics and therapeutic sensitivity. OBJECTIVE: To evaluate whether data derived from routine clinical DNA sequencing of tumors could complement conventional approaches to enable improved diagnostic accuracy. DESIGN, SETTING, AND PARTICIPANTS: A machine learning approach was developed to predict tumor type from targeted panel DNA sequence data obtained at the point of care, incorporating both discrete molecular alterations and inferred features such as mutational signatures. This algorithm was trained on 7791 tumors representing 22 cancer types selected from a prospectively sequenced cohort of patients with advanced cancer. RESULTS: The correct tumor type was predicted for 5748 of the 7791 patients (73.8%) in the training set as well as 8623 of 11 644 patients (74.1%) in an independent cohort. Predictions were assigned probabilities that reflected empirical accuracy, with 3388 cases (43.5%) representing high-confidence predictions (>95% probability). Informative molecular features and feature categories varied widely by tumor type. Genomic analysis of plasma cell-free DNA yielded accurate predictions in 45 of 60 cases (75.0%), suggesting that this approach may be applied in diverse clinical settings including as an adjunct to cancer screening. Likely tissues of origin were predicted from targeted tumor sequencing in 95 of 141 patients (67.4%) with cancers of unknown primary site. Applying this method prospectively to patients under active care enabled genome-directed reassessment of diagnosis in 2 patients initially presumed to have metastatic breast cancer, leading to the selection of more appropriate treatments, which elicited clinical responses. CONCLUSIONS AND RELEVANCE: These results suggest that the application of artificial intelligence to predict tissue of origin in oncologic practice can act as a useful complement to conventional histologic review to provide integrated pathologic diagnoses, often with important therapeutic implications.

A phase II trial of erlotinib in combination with bevacizumab in patients with metastatic breast cancer.

PURPOSE: To evaluate the efficacy and toxicity of erlotinib plus bevacizumab in patients with metastatic breast cancer (MBC), targeting the epidermal growth factor receptor (EGFR/HER1) and the vascular endothelial growth factor (VEGF) pathway. EXPERIMENTAL DESIGN: Thirty-eight patients with MBC were enrolled and treated at two institutions with erlotinib, a small molecule EGFR tyrosine kinase inhibitor (150 mg p.o. daily) plus bevacizumab, an anti-VEGF antibody (15 mg/kg i.v. every 3 weeks). Patients had one to two prior chemotherapy regimens for metastatic disease. The primary end point was response rate by Response Evaluation Criteria in Solid Tumors criteria using a Simon 2-stage design. Secondary end points included toxicity, time to progression, response duration, and stabilization of disease of > or = 26 weeks. Correlative studies were done on tumor tissue, including EGFR expression and mutation analysis. RESULTS: One patient achieved a partial response for 52+ months. Fifteen patients had stable disease at first evaluation at 9 weeks; 4 of these patients had stable disease beyond 26 weeks. Median time to progression was 11 weeks (95% confidence interval, 8-18 weeks). Diarrhea of any grade was observed in 84% of patients (grade 3 in 3%); 76% experienced grade 1 or 2 skin rash, and 18% developed hypertension (grade 3 in 11%). The level of EGFR expression was not predictive of response to therapy. CONCLUSIONS: The combination of erlotinib and bevacizumab was well-tolerated but had limited activity in unselected patients with previously treated MBC. Biomarkers are needed to identify those MBC patients likely to respond to anti-EGFR/HER1 plus anti-VEGF therapy.

Immunization of high-risk breast cancer patients with clustered sTn-KLH conjugate plus the immunologic adjuvant QS-21.

PURPOSE: To determine the clinical toxicities and antibody response against sTn and tumor cells expressing sTn following immunization of high-risk breast cancer patients with clustered sTn-KLH [sTn(c)-KLH] conjugate plus QS-21. EXPERIMENTAL DESIGN: Twenty-seven patients with no evidence of disease and with a history of either stage IV no evidence of disease, rising tumor markers, stage II (>or=4 positive axillary nodes), or stage III disease received a total of five injections each during weeks 1, 2, 3, 7, and 19. Immunizations consisted of sTn(c)-KLH conjugate containing 30, 10, 3, or 1 microg sTn(c) plus 100 microg QS-21. Induction of IgM and IgG antibodies against synthetic sTn(c) and natural sTn on ovine submaxillary mucin were measured before and after therapy. Fluorescence-activated cell sorting analyses assessed reactivity of antibodies to LSC and MCF-7 tumor cells. RESULTS: The most common toxicities were transient local skin reactions at the injection site and mild flu-like symptoms. All patients developed significant IgM and IgG antibody titers against sTn(c). Antibody titers against ovine submaxillary mucin were usually of lower titers. IgM reactivity with LSC tumor cells was observed in 21 patients and with MCF-7 cells in 13 patients. There was minimal IgG reactivity with LSC cells. CONCLUSION: Immunization with sTn(c)-KLH conjugate plus QS-21 is well tolerated and immunogenic in high-risk breast cancer patients. Future trials will incorporate sTn(c) as a component of a multiple antigen vaccine.

A phase II open-label study of ganetespib, a novel heat shock protein 90 inhibitor for patients with metastatic breast cancer.

BACKGROUND: Ganetespib is a small molecule, nongeldanamycin HSP90 inhibitor with potent inhibitory effects on HSP90-dependent oncoproteins of relevance to breast cancer pathogenesis. We therefore tested ganetespib in an unselected cohort of patients with MBC. PATIENTS AND METHODS: Patients were treated with single agent ganetespib at 200 mg/m(2) once weekly for 3 weeks, on a 28-day cycle. Therapy was continued until disease progression. The primary end point was ORR using Reponse Evaluation Criteria in Solid Tumors version 1.1. RESULTS: Twenty-two patients were enrolled with a median age of 51(range, 38-70) years and a median Eastern Cooperative Oncology Group performance status of 0 (range, 0-1). Most patients had at least 2 previous lines of chemotherapy in the metastatic setting. Most common toxicities, largely grade 1/2, were diarrhea, fatigue, nausea, and hypersensitivity reaction. The ORR in this unselected population was 9%, with all responses coming from the subset of patients with HER2-positive MBC (2/13; 15%). One patient with TNBC had objective tumor regression in the lung metastases. The clinical benefit rate (complete response + partial response + stable disease > 6 months) was 9%, median progression-free survival was 7 weeks (95% confidence interval [CI], 7-19), and median overall survival was 46 weeks (95% CI, 27-not applicable). CONCLUSION: The study did not meet the prespecified criteria for ORR in the first stage of the Simon 2-stage model in this heavily pretreated unselected population of MBC. However, activity was observed in trastuzumab-refractory HER2-positive and TNBC. Ganetespib was well tolerated and responses in more targeted populations harboring specific HSP90-dependent oncoproteins justifies its further study, particularly as part of rational combinations.

The joys and challenges of caring for patients.

Optimal care of patients involves the integration of both the scientific and humanistic aspects of medicine. However, the tremendous focus on technology can easily overshadow the personal effect of patient care. The complex relationship between the physician and the patient is a reciprocal one. Not only does the physician influence the experience of the patient, but the patient may leave a significant impression on the physician. Their interactions provide a myriad of opportunities for greater insight into the human condition, but may also contribute toward the struggle of physicians to maintain their own well-being. Enhanced awareness of the significance of these human interactions is at the core of caring for patients.

Encountering grief in patient care.

Grief is essentially unavoidable and is a normal reaction to loss. Grief may be experienced by patients and their loved ones as well as by physicians and members of the health care team in response to the consequences of illness or death. Grief is typified by certain indicators that may significantly effect one's emotional and physical well-being. Although these indicators tend to follow a general pattern, there is variability among individuals. Complicated grief may require psychiatric intervention. Caring for the seriously ill or dying patient may be particularly challenging from an emotional level and may increase the risk of burnout. Recognition of these emotions is a critical aspect of providing compassionate care on a sustainable level. Various strategies may be beneficial in coping with grief, and the exploration of grief may provide greater insight into the humanistic basis of medicine.

The oncologist as the patient with cancer or relative.

To an extent, physicians are familiar with the consequences of illness through their interactions with patients. However, when cancer becomes personal, the physician has an opportunity to gain greater insight into the intricacies of medical care, including its humanistic elements. Physicians who encounter cancer in themselves or in a relative may deepen their understanding of the patient experience. Their views provide a unique perspective, on the basis of the convergence of their medical knowledge and personal reaction to illness. They also confront distinct challenges specific to their work environment. An enhanced recognition of their viewpoints provides valuable information in the quest to alleviate patient suffering and explore the fundamentals of patient care.

Dose dense cyclophosphamide, methotrexate, fluorouracil is feasible at 14-day intervals: a pilot study of every-14-day dosing as adjuvant therapy for breast cancer.

PURPOSE: Cyclophosphamide/methotrexate/fluorouracil (CMF) is a proven adjuvant option for patients with early-stage breast cancer. Randomized trials with other regimens demonstrate that dose-dense (DD) scheduling can offer greater efficacy. We investigated the feasibility of administering CMF using a DD schedule. PATIENTS AND METHODS: Thirty-eight patients with early-stage breast cancer were accrued from March 2008 through June 2008. They were treated every 14 days with C 600, M 40, F 600 (all mg/m2) with PEG-filgrastim (Neulasta®) support on day 2 of each cycle. The primary endpoint was tolerability using a Simon's 2-stage optimal design. The design would effectively discriminate between true tolerability (as protocol-defined) rates of ≤ 60% and ≥ 80%. RESULTS: The median age was 52-years-old (range, 38-78 years of age). Twenty-nine of the 38 patients completed 8 cycles of CMF at 14-day intervals. CONCLUSION: Dose-dense adjuvant CMF is tolerable and feasible at 14-day intervals with PEG-filgrastim support.

Reevaluation of the cellular immune response in breast cancer patients vaccinated with MUC1.

Conjugation of antigens to a carrier protein like keyhole limpet hemocyanin (KLH) has proven effective in clinical trials for inducing antibodies against selected tumor antigens. The impact of this approach on T-cell immunity has not been previously tested. We utilized peripheral blood mononuclear cells (PBMC) obtained at leukapheresis from 6 breast cancer patients vaccinated 4 times each with a 106-amino acid-long MUC1 peptide conjugated with KLH plus immune adjuvant QS-21. Proliferation after 6 days of in vitro culture and an interferon gamma ELISPOT assay with and without 6 days of in vitro sensitization with the immunizing antigen were used. Parallel experiments employed the use of the cytokine IL2. Our results indicate that despite a high response to KLH in all patients with precursor frequencies as high as 1/120 peripheral blood lymphocytes and augmentation of proliferation in excess of 200-fold after vaccination, the T-cell response against MUC1 peptide was minimal and inconsistent. The strength and consistency of the vaccine-induced T-cell response against KLH in these patients excludes general immune incompetence and assay insensitivity or inconsistency as explanations for the weak and inconsistent response against MUC1. We conclude that for any report of augmented T-cell responses against MUC1 to be convincing, one or more postimmunization blood samples will be needed to demonstrate augmented MUC1-specific immunity consistently on multiple occasions. Assuming this criteria, convincing induction of T-cell immunity against MUC1 by vaccination has yet to be described.