P1-06-23: Changes in Gene Expression after One Dose of Trastuzumab (T) in HER2+ Breast Cancer Cell Lines Predict Novel Pathways of Response in HER2 Positive Early Stage Breast Cancer

Introduction: Trastuzumab is a targeted therapy against the HER2 cell surface receptor and has greatly improved prognosis for HER2+ breast cancer patients. Despite decades of research, the mechanism of action of T remains unclear and mechanisms of resistance have not been adequately defined. We sought to determine if models of response and resistance in vitro could identify biomarkers in vivo in HER2+ early breast cancer.

Methods: BT474 and UACC812 cell lines were treated with 10ug/ml T for 0 and 24 hours. Fresh tumor core biopsies were taken at a 2 week timepoint after a single dose of T (8mg/m2) from 80 HER2−overexpressing, early breast cancer patients enrolled on a clinical trial of T>T+C. Nucleic acids were extracted using Qiagen AllPrep and were analyzed with Illumina HT12v3 Beadchip arrays. All arrays were processed at the Yale Center for Genome Analysis (West Haven, CT). Clinical response at surgery was defined as pathologic complete response (pCR), objective response (CR+PR=OR) and non-response (SD+PD=NOR) by RECIST criteria. Gene expression was analyzed in Bioconductor using LIMMA analysis. Pathway analysis was performed using the DAVID bioinformatics resource.

Results: We identified gene expression signatures of T response and resistance in HER2+ breast cancer cell lines across 24-hour exposure to trastuzumab. 180-genes changed significantly in a T-sensitive HER2+ cell line (BT474) and 58 genes changed significantly across treatment in a T-resistant HER2+ cell line (UACC812). We applied these signatures to gene expression profiles from early stage HER2+ breast tumors treated with a single dose of T. The BT474 T-responsive signature was enriched among the changes in expression across treatment for responsive patients. The expression change for the BT474 signature genes was also able to partially cluster responsive and resistant breast tumors by outcome. A subset of the UACC812 T-resistance signature was also enriched in the differential between responsive and resistant tumors prior to T treatment.

Pathway analysis based on the direction of change of genes in pCR and sensitive (BT474) cell lines found gland development/differentiation (enrichment score=4.5), DNA synthesis (ES=4.0), chaperone (ES=2.8) and transcriptional machinery (ES=2.2) to be coordinately downregulated in both sensitive cell lines and tumors. This suggests that the downregulation of differentiation pathways seen in our ‘pCR signature’ (Harris etal: AACR 2011) is not an epiphenomenon of cell loss. There were no significant pathways upregulated in both sensitive tumors and cell lines, however discordant genes were enriched in chromatin regulation pathways (ES=4.1). Of note, our previous findings of amplicon gene downregulation in pCR tumors, yet upregulation in cell lines points to a novel mechanism of chromatin modulation heretofor undiscovered in response to T.

Conclusions: These results demonstrate the value of iterative study of in vitro and in vivo response mechanisms in HER2 cell lines and tumors, and the importance of brief exposure studies in understanding the mechanism of response to T, and other targeted therapies.

Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P1-06-23.

Abstract P3-14-17: Trastuzumab (H) and Rapamycin (R) for Treatment of HER-2 Overexpressing Metastatic Breast Cancer (BC) with Prior Disease Progression on H Based Therapy: Safety and Pharmacodynamic (PD) Results

Background: Preclinical data shows upregulation of the PI3K/AKT pathway and synergistic cytotoxicty between H+R in sensitive and resistant HER-2 overexpressing BC cell lines.

Methods: We designed a phase II clinical trial with an early stopping rule for excess toxicity in the first 4 wks of H+R therapy, defined as ≥3 cases of drug related grade 3 or 4 toxicity observed in the first 9 pts enrolled on trial. We report safety and PD results in 11 pts treated with oral R 6 mg daily + standard dosesand schedule of H. LVEF by ECHO or MUGA was assessed at baseline and every 3 months thereafter; treatment was held for LVEF drop below normal limits or ≥20 percentage points. Pre and post therapy tumor samples were collected when feasible. Circulating tumor (CTC) and endothelial (CEC) cells were collected and analyzed using the Veridex Cell Search System at baseline, wks 1, 2 and 4 and with every restaging.

Results: 11 pts with median age 56 y (range 38-70) treated with H+R were evaluable for safety. Median # cycles was 3 (range 1+ to 12). Non-hematologic grade 3 toxicities in first 4 wks occurred in 3 of the first 9 pts: 1 pt had syncope associated with leucopenia/neutropenia and urinary infection; 1pt had mucositis which responded to dose reduction; 1 pt with known diabetes had grade 3 hyperglycemia, however, this was non-fasting. One pt had grade 3 leucopenia/neutropenia which responded to dose reduction. Grade 3 toxicities for 11 pts in all cycles were as follows: leucopenia/neutropenia (2 pts), non-fasting hyperglycemia (2 pts), syncope (1 pt), hypokalemia (1pt), hyponatremia (1 pt), mucositis (1 pt), rash (2 pts), nail changes (1 pt), thrombosis (1 pt). Infection (urine in 1 pt, skin in 1 pt), anemia (1pt), and elevated PTT (1 pt). One grade 4 non-fasting hyperglycemia occurred in a diabetic pt. LVEF dropped below normal limits and recovered subsequently in 2 pts. Nine pts had baseline and at least one subsequent evaluation for radiographic response by RECIST. Best response was unconfirmed partial response (PR) in 1 pt (8 cycles), stable disease (SD) in 5 pts (12, 6+,6, 3, and 2 cycles) and progression (PD) in 3pts. We combined analysis of AKT pathway markers from 12 baseline and 9 post therapy tissue samples collected on this trial and a phase I study of R +chemotherapy. There were high correlations between baseline AKT pathway markers (mTOR, PTEN, Akt, pAkt, S6K1 and pS6K1; (Spearman's rho-0.401-0.907). This correlation was lost in post therapy samples. Higher mean mTOR levels were seen in non-responders (PD) vs. responders (SD/PR) [p value=0.04]. Eight pts had CTC and CEC results, and were evaluable for response. Baseline CTC levels of >5 cells/7.5 ml were seen in 2/2 pts with PD and 0/6 pts with SD/PR. There was no significant difference in CEC levels between responders and non-responders. Conclusion: R 6 mg daily appears to be well tolerated when combined with standard doses of H. PR in 1 pt and SD for ≥6 months was observed in 2 pts with prior progression on H based therapy. Baseline elevated mTOR level correlates with poor response to R based combinations. Post therapy change in marker expression suggests biologic effect of R.

Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr P3-14-17.

Genome-Wide Profiling of Archived Material from CALGB 9840 and 9342 for Paclitaxel (P) and Trastuzumab (T) Response Biomarkers Using Gene Expression and Copy Number Analysis

Background:Emerging data suggest that RNA obtained from formalin-fixed, paraffin-embedded (FFPE) tissue can produce reliable gene expression profiles. Archived material from two taxane monotherapy studies, CALGB 9342 (comparison of 3 doses of P) and 9840 (weekly vs. q3 week P)were profiled for gene expression and DNA copy number.Methods: A total of 238 patients had primary tumor blocks available from a combined sample size of 680. DNA and RNA was extracted from 1.5mm punch cores and the Ambion Recover-All kit™. quality was measured by spectrophotometric analysis, Bioanalyzer RNA Integrity Number (RIN), and housekeeping genes (RPL13A and Actin). A custom DASL™ array containing 779 genes in two-fold redundancy was designed with genes selected to represent the PAM50 intrinsic subtypes, the Oncotype Dx Score, the Netherlands prognostic signature and the genes most frequently found on recurrent breast cancer amplicons. Several methods for identifying outliers were evaluated, including principal components analysis, pairwise correlations as well as the reproducibility of the platform based on replicate samples.Results: Adequate RNA was obtained from 237/238 of these cases which ranged in age from 12-18 years. Of these, 215/237DASL arrays passed further quality control measures. Adequate DNA for CGH was obtained from 227/238 samples. Analysis of PAM50 intrinsic subtypes showed an excess of basal-like tumors (30%) in the primary tumors of this metastatic cohort compared with expected frequency in an early stage population. Luminal A tumors were less frequent than expected (20%). Patients with basal-like tumors did far worse than other tumor types for both PFS on P (p=0.015) and OS (p=2.7X10-6), which persisted in multivariable analysis (p=0.0047), however the interaction term was not significant (Wald p=0.26). While basal-like tumors had similar PFS and OS on both weekly and q3 week P, luminal A tumors appear to achieve more benefit from weekly P (p= 0.0041).The HER2-enriched expression subtype had a similar prognosis to Luminal A and B tumors. This appeared to be due to the presence of T, as the addition of this agent improved PFS (p=0.026) and OS (p=2.0X10-4). Of note, some centrally confirmed HER2 FISH amplified tumors were classified into luminal A, B, and basal-like subtypes. These tumors have similar prognoses to the overall group, for example the basal-like and HER2 tumors had a poor prognosis despite T (p=0.00086). This suggests that HER2 FISH positive tumors may behave based on the underlying tumor subtype. Sawtooth genomes (45% vs 15%) were more frequent than predicted by an early stage tumor dataset as were simplex genomes (3% vs 24%). HER2 by FISH and CGH were highly concordant suggesting data on gene amplification from this platform is robust.Conclusions: Gene expression and copy number profiling of FFPE material from archived tumor blocks (>10 years) produces quality data for biomarker discovery in CALGB clinical trial datasets. These tools allow discovery of novel patterns of gene expression and genomic aberrations that are associated with differential response to P and T. Further studies using these platforms should be performed.

Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 4032.

Molecular Classification of Normal and Cancer Mammospheres

Introduction: For decades, scientists have attempted to develop models for growing and studying primary tumors outside of the human body. These techniques have ranged from growth assays in immuno-compromised mice to simple monolayer cultures. Unfortunately, these techniques are either too complex and expensive (animal models), or too simple (monolayer cultures) to effectively recapitulate the growth characteristics of human tumors. The recent discovery of cancer stem cells (CSCs) has provided new insight into the reasons that ex-vivo cultures have been unable to provide an accurate model for in-vivo tumors.Hypothesis: Our hypothesis is that three-dimensional mammosphere culture represents a practical means of enriching cancer stem cells and assessing, ex-vivo, the growth characteristics of human breast cancer. We believe this technique will provide a more accurate model system for testing novel drug targets specific for cancer stem cells and tailor specific therapies for patients.Method: By combining ideas from a number of previously described methods, we have recently developed a protocol to grow and maintain breast cancer specimens in a simple three-dimensional culture system using serum-free media, MEGM supplemented with BT-20, insulin, EGF and bFGF. We have successfully cultured both normal and tumor cells from clinical samples as small as a fine needle aspiration. In each case, small subset of epithelial cells from each sample grow into mammospheres.Results: Our preliminary data in characterization of the mammospheres shows that different morphologies and histotypes can be achieved in culture. We are currently using protein expression and transcriptional profiling methods to characterize the cell types within mammospheres and the histotypes of cancer that can generate mammospheres. Using immunofluorescence of a few key protein markers we have found that mammospheres, derived both from normal breast tissue and from tumor tissue samples, express putative stem cells markers as determined by expression of CD44 in the absence of CD24. Only a subset (30-50%) of CD44 positive cells in mammospheres are also positive for ALDH1 in both normal and cancer derived cultures suggesting an undifferentiating population of cells, potentially indicative of the presence of stem cells. Cells in the mammospheres are also positive for both epithelial and myoepithelial markers (e.g. broad-range cytokeratin, CK8, CK5, CK17, vimentin, P63) suggesting variable differentiation potential of progenitor cells. Preliminary transcriptional profiling data suggests that gene sets useful in molecular subtyping of primary tumors may also be informative in mammospheres, and that segregation of profiles of stromal and tumor cells can be elucidated through studies of mammospheres vs. core biopsies.Conclusion: This model system has the potential to be an inexpensive, accurate and robust model for breast cancer. It could help in the understanding of mechanisms of drug resistance, recurrence and metastasis, as well as in predicting treatment success or failure as we move towards personalized medicine.

Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 501.

A randomized controlled clinical trial to determine the optimum duration of G-CSF priming prior to BM stem cell harvesting

BACKGROUND

Harvesting of hemopoietic stem cells (HSC) from G-CSF-primed BM for autologous transplantation is an alternative to collection of unprimed BM or G-CSF-primed peripheral blood (PB). However, the optimum number of days of G-CSF administration for this purpose is unknown. We set out to determine whether cell yields could be optimized by varying the number of days of G-CSF administration prior to BM stem cell harvesting.

METHODS

We conducted a randomized controlled single-center trial of 6 days (the standard) vs. 4 days of G-CSF administration and compared yields of total nucleated cells (TNC), CD34(+) HSC and CFU-GM cells per kilogram patient body weight. Statistical analysis was by Student's t-test.

RESULTS

Twenty-four patients were enrolled; 13 received 6 days and 11 received 4 days of G-CSF administration. Analysis of the first harvest aspirate showed higher proportions of CD34(+) HSC (P=0.02) and CFU-GM (P=0.03) in the 4-day group. For the 6-day and 4-day groups, respectively, the median yield of TNC/kg was 6.5 x 10(8) and 5.4 x 10(8) (P=0.28), of CD34(+) cells/kg 0.56 x 10(6) and 0.98 x 10(6) (P=0.04) and of CFU-GM cells/kg 1.66 x 10(5) and 1.55 x 10(5) (P=0.75).

DISCUSSION

These results suggest that by 6 days the HSC-stimulating effect of G-CSF has passed its peak and that 4 days should be adopted as the standard for G-CSF priming prior to BM stem cell harvesting for autologous transplantation.

Cardiac function following combination therapy with paclitaxel and doxorubicin: an analysis of 657 women with advanced breast cancer

BACKGROUND

To determine the cardiotoxicity of paclitaxel (T) plus doxorubicin (A) combination therapy in women with advanced breast cancer. To define a dose range of A for use in AT.

PATIENTS AND METHODS

The effect of cumulative A dose on risk of congestive heart failure (CHF) and alterations of myocardial contractility (left ventricular ejection fraction [LVEF] decrease > or = 20% or to <50%) was estimated from pooled data from 10 trials of AT.

RESULTS

Thirty-one of 657 patients (4.7%) developed CHF at a median of 6.6 months (range 0.3-24.6) after initiation of AT. CHF was stabilized in 29 patients at a median of 17.3 months after diagnosis (range 4.1-31.2 months). The risk of developing CHF was < or = 5% at a total A dose < or = 380 mg/m2. In patients who received a total A dose > 440 mg/m2, the incidence of CHF was >25% but similar to that of A monotherapy. The risk of CHF was similar in women receiving AT or A monotherapy at a dose < or = 380 mg/m2 (2%-3%). LVEF progressively decreased in patients who received AT, especially at a cumulative A dose > 380 mg/m2. LVEF decreases were more frequent in patients who later developed CHF, but the majority of CHF patients did not experience LVEF alterations prior to symptoms. LVEF recovered after discontinuation of A in 25 of 67 women who developed LVEF < 50%.

CONCLUSION

The reported cardiac effects are consistent with anthracycline-related cardiotoxicity. AT is associated with a cardiac risk similar to that of A monotherapy up to a cumulative A dose of 340-380 mg/m2.

Phase II trial with dose titration of paclitaxel for the therapy of human immunodeficiency virus-associated Kaposi's sarcoma

PURPOSE

To investigate the antitumor activity and safety of paclitaxel in patients with advanced human immunodeficiency virus (HIV)-associated Kaposi's sarcoma (KS).

PATIENTS AND METHODS

Twenty-nine patients with advanced HIV-associated KS were enrolled. The patients were overall quite immunosuppressed (median CD4 count, 15 cells/microL). Paclitaxel was initially administered at 135 mg/m2 over 3 hours every 3 weeks without filgrastim support; the dose was increased as tolerated to a maximum of 175 mg/m2. Patients who failed to respond or progressed could then receive filgrastim support or paclitaxel administered over 96 hours.

RESULTS

Of 28 assessable patients, 20 had major responses (18 partial responses [PRs], one clinical complete response [CR], and one CR), for a major response rate of 71.4% (95% confidence interval [CI], 51.3% to 86.8%). Each of the five patients with pulmonary KS responded, as did all four assessable patients who had previously received anthracycline therapy for KS. Of six patients who went on to receive a 96-hour infusion of paclitaxel, five had major responses. Neutropenia was the most frequent dose-limiting toxicity; possible novel toxicities included late fevers, late rash, and eosinophilia. Two patients developed an elevated creatinine concentration and one cardiomyopathy.

CONCLUSION

Paclitaxel has substantial activity against advanced HIV-associated KS as a single agent, even in patients with pulmonary involvement or who had previously received anthracyclines. Further research is needed to define the optimal treatment schedule and its role vis-a-vis the other available therapies for this disease.

A system for automated procedure documentation

Quantitative information regarding competency in performing procedures is invaluable in attesting to a house officer's abilities when assessments need to be made. Thus, the means by which procedures, proficiency, and experience can be recorded and organized in an automated fashion need not be any further away than the nearest micro-computer. The procedure documenting system (PDS) automates the documenting of procedures and quantitative information pertaining to these procedures, and shortens the time demanded of the house officer to document procedures from a mean of 5.5 to 1.9 min. This paper is a description of the work done to automate the recording of procedures performed by housestaff at a teaching hospital and the acceptance by housestaff. This computer program takes the specific information and places it into a specialized database so that house officers can have detailed documentation attesting to their proficiency in various procedures. In addition, the program director can access the database to document the quantity of procedures to which an individual house officer has attained competence.

Pancytopenia in allogeneic marrow transplant recipients: role of cytomegalovirus

We describe the clinical course of three cytomegalovirus-antibody-positive allogeneic marrow graft recipients who developed progressive pancytopenia during the third month post-transplant. Bone marrow biopsy cores were hypocellular without evidence of disease recurrence. Haemopoietic progenitor assays demonstrated a decrease of all assayable progenitors. Cytomegalovirus was identified from the blood and urine of all three patients. However, none of the patients developed other manifestations of cytomegalovirus infection such as pneumonitis, hepatitis and enteritis. The therapeutic use of ganciclovir and intravenous immunoglobulins resulted in prompt resolution of both viraemia and viruria in all three patients, and of pancytopenia in two patients.

Expression of membrane-bound burst-promoting activity is mediated by allogeneic effector cells

To investigate whether "self" and "non-self" recognition processes are involved in murine erythropoiesis, the expression of membrane-bound burst-promoting activity (mBPA) was determined for B lymphocytes purified from spleens of CF-1, C57 BL/6J, B6021-7115, and CAF-1J mice using syngeneic and allogeneic bone marrow cultures. Addition of B lymphocyte conditioned medium (LCM), shed membrane-derived vesicles, or intact plasma membranes prepared from syngeneic murine cells stimulated erythroid burst-forming unit (BFU-E) proliferation by two- to three-fold above control levels. BFU-E proliferation was increased by six- to eight-fold, however, when LCM, shed membrane vesicles, or plasma membranes purified from allogenic B lymphocytes were used as sources of growth-stimulatory activity. Bioactivity was immunoprecipitated from detergent extracts of membranes purified from both allogeneic and syngeneic lymphocytes with a monoclonal antibody that specifically recognizes mBPA, suggesting that the factors expressed by these cells share antigenic determinants. The results indicate that allogeneic effector cells are a more potent source of mBPA-like molecules than are syngeneic cells, suggesting that immune mechanisms may be involved in inducing erythroid growth factor expression at the B cell surface.

Factors affecting recovery after peripheral blood stem cell transplantation

To determine the factors that might be involved in successful engraftment following administration of autologous peripheral blood stem cells (PBSC) obtained by leukapheresis to replace bone marrow after ablation therapy for malignant disease, four patients were examined in detail. One who had had pelvic radiation as well as chemotherapy had a prolonged, gradual but complete recovery. One who received granulocyte-macrophage-colony stimulating factor (GM-CSF) following PBSC infusion showed a rapid recovery phase, but developed high fever (culture negative) associated with arrested hematopoiesis and died with central nervous system (CNS) symptoms after 120 days. Two other patients recovered without major incident. Concentration of PBSC was analyzed by: (1) approximation by counting mononuclear cells (MNC); (2) CD34 cells by flow cytometric analysis; and (3) colony forming units-granulocyte macrophage (CFU-GM) by colony formation in microtiter plates. The time to BM recovery (retics > 1.0 percent, neutrophils > 500 per cumm, platelets > 50,000 per cumm) was determined by following daily counts. Engraftment appeared to depend upon an adequate minimum dose of PBSC, but ultimate recovery of the patient seemed to be determined by ancillary factors, especially CNS infection. These patients illustrate that while rapidity of bone marrow (BM) recovery may be related to PBSC dose, other factors, particularly infection, influence patient outcome.

Immunologic enhancement of B16 melanoma growth

In certain experimental tumor models, tumor growth is less pronounced in immune deficient animals. Characteristically, tumors such as the murine B16 melanoma and Lewis lung carcinoma (3LL) are weakly antigenic. We proposed that with such tumors that are weakly antigenic, growth is enhanced by T-cell factors. Young mice were inoculated with irradiated B16 cells in complete Freunds adjuvant (CFA) on three occasions, each separated by 2 weeks. Specific antibody (IgG) to B16 membrane antigens was detected by an enzyme-linked immunosorbent assay (ELISA) after the first injection, and it continued to rise for 6 weeks. B16 growth was compared in 20 mice that had received irradiated B16 in CFA or CFA alone by the same schedule previously. Despite the previous sensitization, the rates of tumor appearance and growth were similar. In an additional experiment involving 23 mice that had received B16 immunization, the period of time in which a palpable tumor developed after the injection of viable B16 cells did not correlate with anti-B16 antibody level. It appeared that detectable antibody to B16 antigens was of little consequence. To explain why B16 primary growth and metastases were reduced in immune deficient hosts, we proposed that lymphocytes might enhance tumor growth. To demonstrate this, splenic lymphocytes from tumor-bearing (B16 or 3LL) or control mice were injected with B16 cells into young, immune competent hosts. Tumors (B16) developed earlier and growth was more rapid in mice that received spleen cells from tumor-bearing (B16) mice. Subsequent cell depletion experiments to determine the mediator of tumor enhancement implicated a T-cell fraction that was neither of T-helper nor T-suppressor cell type phenotypically. Immune deficiency states that are associated with dysfunction of those cells that account for tumor enhancement might explain the reduced tumor aggressiveness that is observed frequently in these conditions.

Relationship between plasma level and therapeutic effect of nortriptyline

The relationship between plasma concentration of nortriptyline and therapeutic effect after two weeks' treatment with the drug was investigated in 29 psychiatric inpatients. Endogenous depression was diagnosed in all patients. Amelioration of depressive symptoms was estimated as reduction in score on a rating scale, based on a psychiatric interview. Amelioration was not correlated to the patient's sex or age. There was a curved relationship between plasma level of nortriptyline and therapeutic effect. Amelioration was most pronounced in the intermediate plasma level range (50-139 ng nortriptyline/ml plasma) and was slight both at lower and at higher plasma levels. This type of relationship may be due to the dual action of tricyclic antidepressants which has been found in animal experiments. On larger dosages a phenothiazine-like blockade of the monoaminergic receptor is added to the blockade of monoamine reuptake thought to be related to the antidepressant action of the drugs.THIS STUDY THUS SUGGESTS TWO POSSIBLE REASONS FOR A THERAPEUTIC FAILURE WITH NORTRIPTYLINE: a too low or a too high plasma level. The large individual variation in the pharmacokinetics of the tricyclic antidepressants makes prediction of plasma level from dosage in a given individual virtually impossible without knowledge of rate of elimination and apparent volume of distribution. Hence monitoring plasma levels may be a way to increase the efficacy of treatment with these drugs.

Correlation of subjective side effects with plasma concentrations of nortriptyline

Plasma levels of tricyclic antidepressant drugs vary considerably between individuals receiving the same amount of drug. The bearing of this variation on the occurrence of subjective side effects was investigated in 40 psychiatric inpatients with depressive disorders. Plasma levels were determined before and during four weeks of treatment with nortriptyline 50 mg. three times a day and patients were rated for subjective side effects, the assessors being unaware of the plasma levels of the drug.Plasma levels varied widely between individual patients, but in any given patient the plasma level tended to be constant over a period of time. The side effects of nortriptyline diminished significantly with time and were in most cases absent during the fourth week of treatment. There was a significant positive correlation between plasma level of nortriptyline and subjective side effects.The steady-state plasma level of a drug which is metabolized is usually a more important determinant for its effect than dosage, since it reflects the amount of drug available for biological action. Very high plasma levels of nortriptyline should presumably be avoided, since there is no evidence that they are needed for therapeutic effect and they are potentially harmful.