Job turnover and its correlates among residency program directors in internal medicine: a three-year cohort study.

PURPOSE: In 1983, 43% of internal medicine residency program directors had held their positions for less than three years. The purposes of this study were to determine the job turnover rate for internal medicine program directors, and the characteristics of program directors and residency programs that are associated with job turnover. METHOD: In October 1996, questionnaires were sent to all non-military internal medicine residency program directors in the continental United States listed by the Accreditation Council for Graduate Medical Education (ACGME). The questionnaire covered demographics, program characteristics, and job satisfaction. In October 1999, an updated ACGME list was used to contact programs to verify changes in program directors and determine the dates of change. RESULTS: A total of 262 usable responses were received. At the beginning of the study, 49% of the respondents had been on the job for three years or less, and 74 (29%) were no longer program directors three years later. Overall job satisfaction was highly associated (p <.01) with turnover. Multivariate Cox regression modeling yielded four variables independently associated with turnover: low satisfaction with colleague relationships (hazard ratio = 3.2, 95% CI = 1.6-6.4), a high percentage of administrative work time (HR = 2.9, 95% CI = 1.4-6.2), perceiving the job as a "stepping stone" (HR = 1.8, 95% CI = 1.0-3.2), and having had formal training to deal with problem residents (HR = 0.6, 95% CI = 0.4-1.1). Respondents with burnout, with the titles of program director and chair or department chief, and with less than two years on the job had nonsignificant trends toward job turnover. Variables not associated with turnover included gender, rank, salary, and program size. CONCLUSIONS: Yearly turnover for internal medicine residency program directors is substantial. The four independent predictors of turnover identified in this study should be of interest to institutions recruiting or retaining program directors and to aspiring program directors.

Personal growth in medical faculty: a qualitative study.

BACKGROUND: A physician's effectiveness depends on good communication, and cognitive and technical skills used with wisdom, compassion, and integrity. Attaining the last attributes requires growth in awareness and management of one's feelings, attitudes, beliefs, and life experiences. Yet, little empiric research has been done on physicians' personal growth. OBJECTIVE: To use qualitative methods to understand personal growth in a selected group of medical faculty. DESIGN: Case study, using open-ended survey methods to elicit written descriptions of respondents' personal growth experiences. SETTING: United States and Great Britain. PARTICIPANTS: Facilitators, facilitators-in-training, and members of a personal growth interest group of the American Academy on Physician and Patient, chosen because of their interest, knowledge, and experience in the topic area and their accessibility. MEASUREMENTS: Qualitative analysis of submitted stories included initially identifying and sorting themes, placing themes into categories, applying the categories to the database for verification, and verifying findings by independent reviewers. RESULTS: Of 64 subjects, 32 returned questionnaires containing 42 stories. Respondents and nonrespondents were not significantly different in age, sex, or specialty. The analysis revealed 3 major processes that promoted personal growth: powerful experiences, helping relationships, and introspection. Usually personal growth stories began with a powerful experience or a helping relationship (or both), proceeded to introspection, and ended in a personal growth outcome. Personal growth outcomes included changes in values, goals, or direction; healthier behaviors; improved connectedness with others; improved sense of self; and increased productivity, energy, or creativity. CONCLUSIONS: Powerful experiences, helping relationships, and introspection preceded important personal growth. These findings are consistent with theoretic and empiric adult learning literature and could have implications for medical education and practice. They need to be confirmed in other physician populations.

A job-satisfaction measure for internal medicine residency program directors.

PURPOSE: To develop a job-satisfaction measure that encompasses the multifaceted job of internal medicine residency program directors. METHOD: Questions were devised to measure program directors satisfaction with various facets of their jobs. In 1996, the authors surveyed all non-military internal medicine program directors in the United States. RESULTS: Of the program directors surveyed, 301 (78%) responded. More respondents than non-respondents held the title of department chairperson in addition to the title of program director (22% vs 7%). Factor analysis and correlation analysis yielded a multifaceted measure (termed PD-Sat) composed of 20 questions and six facets (work with residents, colleague relationships, resources, patient care, pay, and promotion) that made sense based on literature review and discussions with program directors (face validity). The PD-Sat had good internal reliability (Cronbach's alpha = .88), as had each of its six facets (Cronbach's alphas = .60-.90). The six facets correlated modestly with one another (Pearson's r2 = .12-.67), suggesting they were measuring different aspects of a common concept. The PD-Sat correlated significantly with an established four-question global job-satisfaction scale used in previous studies (Pearson's r2 = .33) demonstrating concurrent validity. Scores on the PD-Sat predicted whether program directors were considering, seeking, or making a job change (predictive validity). The PD-Sat performed comparably well in subsets of program directors who were and were not department chairs, suggesting that it might be applicable to different populations of program directors. CONCLUSION: The authors have developed a new facet-specific job-satisfaction measure that is reliable and valid for assessing the job satisfaction of internal medicine program directors. Because job descriptions for program directors in other specialties are similar, it may also be useful in these populations.

Attributes of excellent attending-physician role models.

BACKGROUND: Although effective role models are important in medical education, little is known about the characteristics of physicians who serve as excellent clinical role models. We therefore conducted a case-control study to identify attributes that distinguish such physicians from their colleagues. METHODS: We asked members of the internal-medicine house staff at four teaching hospitals to name physicians whom they considered to be excellent role models. A total of 165 physicians named by one or more house-staff members were classified as excellent role models (these served as the case physicians in our study). A questionnaire was sent to them as well as to 246 physicians who had residency-level teaching responsibilities but who were not named (controls). Of these 411 physicians, 341 (83 percent) completed questionnaires while unaware of their case-control status. RESULTS: Of the 341 attending physicians who responded, 144 (42 percent) had been identified as excellent role models. Having greater assigned teaching responsibilities was strongly associated with being identified as an excellent role model. In the multivariate analysis, five attributes were independently associated with being named as an excellent role model: spending more than 25 percent of one's time teaching (odds ratio, 5.12; 95 percent confidence interval, 1.81 to 14.47), spending 25 or more hours per week teaching and conducting rounds when serving as an attending physician (odds ratio, 2.48; 95 percent confidence interval, 1.15 to 5.37), stressing the importance of the doctor-patient relationship in one's teaching (odds ratio, 2.58; 95 percent confidence interval, 1.03 to 6.43), teaching the psychosocial aspects of medicine (odds ratio, 2.31; 95 percent confidence interval, 1.23 to 4.35), and having served as a chief resident (odds ratio, 2.07; 95 percent confidence interval, 1.07 to 3.98). CONCLUSIONS: These data suggest that many of the attributes associated with being an excellent role model are related to skills that can be acquired and to modifiable behavior.

Relation of low-severity violence to women's health.

OBJECTIVE: To determine if women who experience low-severity violence differ in numbers of physical symptoms, psychological distress, or substance abuse from women who have never been abused and from women who experience high-severity violence. DESIGN: Cross-sectional, self-administered, anonymous survey. SETTING: Four community-based, primary care, internal medicine practices. PATIENTS: Survey respondents were 1,931 women aged 18 years or older. SURVEY DESIGN: Survey included questions on violence; a checklist of 22 physical symptoms; the Symptom Checklist-22 (SCL-22) to measure depression, anxiety, somatization, and self-esteem; CAGE questions for alcohol use; and questions about past medical history. Low-severity violence patients had been "pushed or grabbed" or had someone "threaten to hurt them or someone they love" in the year prior to presentation. High-severity violence patients had been hit, slapped, kicked, burned, choked, or threatened or hurt with a weapon. MAIN RESULTS: Of the 1,931 women, 47 met criteria for current low-severity violence without prior abuse, and 79 met criteria for current high-severity violence without prior abuse, and 1,257 had never experienced violence. The remaining patients reported either childhood violence or past adult abuse. When adjusted for socioeconomic characteristics, the number of physical symptoms increased with increasing severity of violence (4.3 for no violence, 5.3 for low-severity violence, 6.4 for high-severity violence, p < .0001). Psychological distress also increased with increasing severity of violence (mean total SCL22 scores 32.6 for no violence, 35.7 for low-severity violence, 39.5 for high-severity violence, p < .0001). Women with any current violence were more likely to have a history of substance abuse (prevalence ratio [PR] 1.8 for low-severity, 1.9 for high-severity violence) and to have a substance-abusing partner (PR 2.4 for both violence groups). CONCLUSIONS: In this study, even low-severity violence was associated with physical and psychological health problems in women. The data suggest a dose-response relation between the severity of violence and the degree of physical and psychological distress.

Analysis of T-cell receptor gene rearrangement in lymph nodes of patients with mycosis fungoides. Prognostic implications.

OBJECTIVE: To determine the prognostic value of analyzing lymph node (LN) DNA from patients with mycosis fungoides for the presence of a monoclonal T-cell population. DESIGN: Inception cohort study. SETTING: A tertiary care referral center in Seattle, Wash. PATIENTS: Fifty-five uniformly staged patients with the diagnosis of mycosis fungoides and who had a lymph node biopsy, 21 with clinically abnormal nodes and 34 with normal nodes. MAIN OUTCOME MEASURES: Lymph nodes were evaluated by Southern blot analysis for T-cell receptor beta-chain (TCRB) gene rearrangement and by histopathologic examination for the LN classification using the National Cancer Institute system. Patients were observed clinically for a mean (+/- SD) of 4.7 +/- 3.4 years. RESULTS: Patients with detectable TCRB gene rearrangement in lymph node DNA had an increased likelihood of a poor clinical outcome and a decreased probability of survival (P < .001 for both) compared with patients with the TCRB germline. Although patients with clinically enlarged nodes were more likely to have the TCRB gene rearranged, those with normal nodes and the TCRB gene rearranged also had a poor clinical outcome and a decreased probability of survival. Similar to those with the TCRB gene rearranged, most patients with advanced histopathologic changes (LN3 and LN4) had a poor prognosis. The presence of a rearranged TCRB gene, however, correctly predicted some patients with intermediate LN scores (LN2) who had a poor clinical outcome. CONCLUSIONS: Detection of a monoclonal T-cell population, as demonstrated by a rearranged TCRB gene on Southern blot analysis, in LNs of patients with mycosis fungoides is predictive of a poor clinical outcome and a reduced probability of survival. Lymph node TCRB gene analysis provides additional prognostic information for patients with mycosis fungoides with intermediate LN histopathology.

Clinical characteristics of women with a history of childhood abuse: unhealed wounds.

OBJECTIVES: To determine the prevalence of childhood physical or sexual abuse in women seen in primary care practices; to identify physical and psychologic problems associated with that abuse; and to compare the effects of childhood physical vs sexual abuse and childhood vs adult abuse. DESIGN: Cross-sectional, self-administered, anonymous survey. SETTING: Four community-based, primary care internal medicine practices. PATIENTS: A total of 1931 women of varied age and marital, educational, and economic status examined from February through July 1993. MAIN OUTCOME MEASURES: Prevalence of physical and sexual abuse, physical symptoms, psychological symptoms (Symptom Checklist-22), alcohol abuse (CAGE questions), and street drug use. RESULTS: Of the 1931 respondents, 424 (22.0%) reported childhood or adolescent physical or sexual abuse. Compared with women who reported never having experienced abuse (n=1257), women who reported abuse as children but not adults (n=204) had more physical symptoms (mean+/-SE, 6.2+/-0.2 vs 4.0+/-0.9; P<.001) and had higher scores for depression, anxiety, somatization, and interpersonal sensitivity (low self-esteem) (P<.001); were more likely to be abusing drugs (prevalence ratio [PR], 4.7; 95% confidence interval [CI], 2.9-7.6) or to have a history of alcohol abuse (PR, 2.2; 95% CI, 1.5-3.2); were more likely to have attempted suicide (PR, 3.7; 95% CI, 2.6-5.1); and were more likely to have had a psychiatric admission (PR, 3.2; 95% CI, 2.2-4.7). Women abused only as children did not differ from women who reported current, but not childhood, abuse in number of physical symptoms, emotional distress, substance abuse, or suicide attempts. Patients who reported both childhood and adult abuse had higher levels of psychological problems and physical symptoms than those who reported childhood or adult abuse alone. CONCLUSIONS: Childhood physical or sexual abuse is associated with adult health problems including physical symptoms, psychological problems, and substance abuse; for many variables, this association is as strong as for patients experiencing current abuse.

The "battering syndrome": prevalence and clinical characteristics of domestic violence in primary care internal medicine practices.

OBJECTIVES: To determine the prevalence of domestic violence among female patients and to identify clinical characteristics that are associated with current domestic violence. DESIGN: Cross-sectional, self-administered, anonymous survey. SETTING: 4 community-based, primary care internal medicine practices. PATIENTS: 1952 female patients of varied age and marital, educational, and economic status who were seen from February to July 1993. MEASUREMENTS: The survey instrument included previously validated questions on physical and sexual abuse, alcohol abuse, and emotional status and questions on demographic characteristics, physical symptoms, use of street drugs and prescribed medications, and medical and psychiatric history. RESULTS: 108 of the 1952 respondents (5.5%) had experienced domestic violence in the year before presentation. Four hundred eighteen (21.4%) had experienced domestic violence sometime in their adult lives, 429 (22.0%) before age 18 years, and 639 (32.7%) as either an adult or child. Compared with women who had not recently experienced domestic violence, currently abused patients were more likely to be younger than 35 years of age (prevalence ratio [PR], 4.1 [95% CI, 2.8 to 6.0]); were more likely to be single, separated, or divorced (PR, 2.5 [CI, 1.7 to 3.6]); were more likely to be receiving medical assistance or to have no insurance (PR, 4.3 [CI, 2.8 to 6.6]); had more physical symptoms (mean, 7.3 +/- 0.38 compared with 4.6 +/- 0.08; P < 0.001); had higher scores on instruments for depression, anxiety, somatization, and interpersonal sensitivity (low self-esteem) (P < 0.001); were more likely to have a partner abusing drugs or alcohol (PR, 6.3 [CI, 4.4 to 9.2]); were more likely to be abusing drugs (PR, 4.4 [CI, 1.9 to 10.4]) or alcohol (PR, 3.1 [CI, 1.5 to 6.5]); and were more likely to have attempted suicide (PR, 4.3 [CI, 2.8 to 6.5]). They visited the emergency department more frequently (PR, 1.7 [CI, 1.2 to 2.5]) but did not have more hospitalizations for psychiatric disorders. In a logistic regression model into which 9 risk factors were entered, the likelihood of current abuse increased with the number of risk factors, from 1.2% when 0 to 1 risk factors were present to 70.4% when 6 to 7 risk factors were present. CONCLUSIONS: In a large, diverse, community-based population of primary care patients, 1 of every 20 women had experienced domestic violence in the previous year; 1 of every 5 had experienced violence in their adult life; and 1 of every 3 had experienced violence as either a child or an adult. Current domestic violence is associated with single or separated status, socioeconomic status, substance abuse, specific psychological symptoms, specific physical symptoms, and the total number of physical symptoms.

PIP: This cross-sectional study determined the prevalence of domestic violence among female patients presenting to four community-based primary care internal medicine practices in Baltimore, Maryland, between February and July, 1993. Furthermore, it identified clinical characteristics associated with domestic violence. A total of 1952 female patients of diverse socioeconomic backgrounds participated in a self-administered, anonymous survey that solicited data on physical and sexual abuse, alcohol abuse, emotional status, demographic characteristics, physical symptoms, use of street drugs and prescribed medications, and medical and psychiatric history. Of the 1952 respondents, 108 (5.5%) had experienced domestic violence in the previous year, 418 (21.4%) had experienced violence sometime in their adult lives, 429 (22%) before age 18 years, and 639 (32.7%) as either an adult or a child. Current violence status is associated with single or separated status, substance abuse, specific psychological symptoms, specific physical symptoms, and the total number of physical symptoms. In a logistic regression model, the likelihood of current abuse increased with the number of risk factors. The magnitude of these associations supports the idea that domestic violence is a significant medical public health problem. Detection of domestic violence by physicians or other health care professionals might alter both the diagnostic and treatment plans for these women.

Improving physicians' interviewing skills and reducing patients' emotional distress. A randomized clinical trial.

BACKGROUND: Despite high prevalence, emotional distress among primary care patients often goes unrecognized during routine medical encounters. OBJECTIVE: To explore the effect of communication-skills training on the process and outcome of care associated with patients' emotional distress. METHODS: A randomized, controlled field trial was conducted with 69 primary care physicians and 648 of their patients. Physicians were randomized to a no-training control group or one of two communication-skills training courses designed to help physicians address patients' emotional distress. The two training courses addressed communication through problem-defining skills or emotion-handling skills. All office visits of study physicians were audiotaped until five emotionally distressed and five nondistressed patients were enrolled based on patient response to the General Health Questionnaire. Physicians were also audiotaped interviewing a simulated patient to evaluate clinical proficiency. Telephone monitoring of distressed patients for utilization of medical services and General Health Questionnaire scores was conducted 2 weeks, 3 months, and 6 months after their audiotaped office visits. RESULTS: Audiotape analysis of actual and simulated patients showed that trained physicians used significantly more problem-defining and emotion-handling skills than did untrained physicians, without increasing the length of the visit. Trained physicians also reported more psychosocial problems, engaged in more strategies for managing emotional problems with actual patients, and scored higher in clinical proficiency with simulated patients. Patients of trained physicians reported reduction in emotional distress for as long as 6 months. CONCLUSIONS: Important changes in physicians' communication skills were evident after an 8-hour program. The training improved the process and outcome of care without lengthening the visits.

Requirements for the generation of a Lyt-2+ T-cell proliferative response to a syngeneic tumor in the absence of L3T4+ T-cells.

Tumors may contain immunogenic antigens that are only recognizable in the context of class I, and not of class II, MHC molecules. Therefore, methods were developed to analyze the capacity of Lyt-2+ T-cells to respond to a syngeneic tumor in the absence of a contribution by L3T4+ T-cells. Conditions were defined in which purified Lyt-2+ T-cell populations, as well as L3T4+ T-cell populations, isolated from immune B6 spleen cells, could be induced to proliferate specifically in response to FBL, a retrovirally induced syngeneic tumor, without the addition of exogenous lymphokines. The purity of the subset responses was documented functionally by selective inhibition of the proliferative response of only the appropriate subset following addition of anti-Kb/Db or anti-I-Ab. The antigen and accessory cell (AC) requirements for triggering immune Lyt-2+ and L3T4+ T-cell populations were examined. The response of L3T4+ populations was predominantly specific for retrovirus envelope gp70, whereas Lyt-2+ populations predominantly recognized tumor antigens other than gp70, consistent with the hypothesis that some tumor antigens may be preferentially recognized by only class I- or class II-restricted T-cells. The FBL-stimulated proliferative response of each T-cell subset was dependent upon the presence of syngeneic AC. However, exogenous interleukin 1 was able to replace AC during the response of Lyt-2+ populations, whereas L3T4+ populations required AC also to biochemically process tumor-derived antigen and present it in the context of class II MHC molecules. The results suggest that under some conditions only the presence of AC or interleukin 1 may be limiting for the induction of antitumor responses by Lyt-2+ populations. These studies analyzed the ability to trigger purified Lyt-2+ T-cells in vitro following in vivo priming to tumor, and it remained possible that L3T4+ T-cells made an essential contribution during in vivo priming. Therefore, L3T4(+)-deficient mice were primed with FBL in vivo, and the Lyt-2+ T-cell response was assessed. Although priming was clearly less efficient in the absence of L3T4+ T-cells, Lyt-2+ T-cells from L3T4(+)-deficient mice proliferated and became cytolytically active following stimulation with FBL. Thus, under appropriate conditions, Lyt-2+ T-cells can generate an effective antitumor response in the absence of L3T4+ T-cells or exogenous lymphokines.

A formula for estimating pretest probability: evaluation and clinical application.

Knowledge of the prevalence (or pretest probability) of a disease is necessary for the interpretation of the results of a diagnostic test in a specific population of patients. This paper evaluates a formula for estimating the prevalence of a disease in a population, based on the proportion of patients with abnormal test results in that population and the known sensitivity and specificity of the test. The authors tested the formula by using it to estimate the prevalence of myocardial infarction in 215 patients with chest pain admitted to a coronary care unit, based on results of initial total creatine kinase determinations. The estimated prevalence was 30%. The true prevalence of myocardial infarction, based on established diagnostic criteria, was 25% (95% confidence interval 19.2%-30.8%). To further evaluate the formula, a sensitivity analysis was performed. Errors in estimated prevalence were inversely related to test sensitivity and specificity, positively related to the magnitude of the differences between presumed and true test sensitivity and specificity, and complexly related to the true prevalence of disease. This formula permits the estimation of prevalence of a disease in a population without resorting to the use of a "gold standard" test, which is often invasive or impractical. Situations are presented where the formula could be used to evaluate and improve the utilization of laboratory tests.

An evaluation of residency training in interviewing skills and the psychosocial domain of medical practice.

Competent use of interviewing skills is important for the care of all patients but is especially critical, and frequently deficient, in meeting the needs of patients experiencing emotional distress. This study presents an evaluation of a curriculum in communication and psychosocial skills taught to first-year medical residents. A randomized experimental design compared trained and untrained residents' (n = 48) performances with a simulated patient presenting with atypical chest pain and psychosocial distress. Evaluation was based on analysis of videotapes, simulated patient report of residents' behaviors, and chart notation. Trained compared with untrained residents asked more open-ended questions and fewer leading questions, summarized main points more frequently, did more psychosocial counseling, and were rated as having better communication skills by the simulated patient. The use of more focused and psychosocially directed questions, and fewer leading and grab-bag questions, was associated with more accurate diagnoses and management recorded in the medical chart. However, no significant difference was found in the charting practices of trained versus untrained residents.

Use of an outpatient medical record audit to achieve educational objectives: changes in residents' performances over six years.

OBJECTIVE: To evaluate the effectiveness of a process whereby a faculty-resident committee annually audits outpatient record keeping and preventive care practices and provides feedback to resident physicians. DESIGN: Pre- and postfeedback audits with interventions and observations repeated over six consecutive academic years. SETTING: The adult primary care practice of housestaff in a university-affiliated hospital. SUBJECTS: All 139 physicians in an internal medicine residency program from 1981-82 through 1986-87, of whom 37 were present for three consecutive years. INTERVENTION: Each year, residents were given individualized, detailed, typewritten feedback based on audits of their outpatient records. MEASUREMENTS AND MAIN RESULTS: Each resident physician had a minimum of four (mean 5.2) outpatient records per year audited against standards for record-keeping practices and the provision of preventive care. Overall performance scores for each resident audit improved from a mean of 39.7 +/- 12.3 (SD) in 1981-82 to a mean of 58.5 +/- 14.1 (SD) in 1986-87 (possible range 0 to 100, observed range 9.4 to 86.6). The overall performance scores of individual residents, who received two cycles of feedback, improved an average of 11.5 (95% confidence limits 7.6, 15.3), from a mean of 48.4 +/- 11.4 (SD) during their first year of residency to 59.8 +/- 13.9 (SD) during their third year. General (primary care) and traditional-track residents improved at similar rates, although mean performance scores were consistently higher for general than for traditional-track residents. Analysis of variance revealed that all changes and differences were statistically significant. CONCLUSIONS: An ongoing chart audit and feedback system can be associated with improvements both in the performance of individual residents and in the long-term performance of a residency program.

Enhancement by interleukin 4 of interleukin 2- or antibody-induced proliferation of lymphocytes from interleukin 2-treated cancer patients.

Systemic interleukin 2 (IL-2) and IL-2-activated lymphocytes have induced tumor regression in some cancer patients. The IL-2-activated cells have usually been generated by obtaining peripheral blood mononuclear cells (PBMC) from cancer patients shortly after systemic IL-2 therapy and culturing them with IL-2 in vitro. In an effort to augment the ex vivo generation of such cells preactivated in vivo, we examined the proliferative responses of PBMC from IL-2-treated cancer patients to several proliferative signals including IL-2, interleukin 4 (IL-4), and mitogenic antibodies to CD3 and CD28. Although much is known about the response of normal PBMC to these signals, the possibility was considered that the response of lymphocytes preactivated by IL-2 in vivo might differ from that of normal PBMC. Accordingly, PBMC obtained from ten normal, healthy controls and from 17 patients with advanced cancer 1 to 3 days after systemic IL-2 therapy were cultured for 4 days with IL-4 (1000 units/ml) and/or IL-2 (10 units/ml or 1000 units/ml) or with combinations of IL-4 and anti-CD3 +/- anti-CD28, and they were then tested for proliferation by [3H]thymidine incorporation. IL-4 failed to induce proliferation of normal PBMC and inhibited IL-2-induced proliferation, whereas IL-4 alone induced proliferation in PBMC from five of 11 IL-2-treated patients and did not inhibit but augmented the proliferation induced by IL-2 (10 units/ml and 1000 units/ml) in PBMC from six of nine patients and five of 11 patients, respectively. Anti-CD3 induced proliferation in PBMC from eight of nine patients, and the proliferation was consistently augmented by coculture with anti-CD28. Finally, IL-4 significantly augmented the proliferative responses of PBMC from IL-2-treated patients to anti-CD3, as well as to the combination of anti-CD3 and anti-CD28. Thus, in PBMC from IL-2-treated cancer patients, IL-4 enhanced the in vitro proliferation induced by IL-2 or by anti-CD3 +/- anti-CD28. The results suggest that IL-4 and/or mitogenic antibodies may be useful in augmenting the ex vivo generation of lymphocytes for clinical adoptive immunotherapy.

Induction of lymphokine-activated killer activity by interleukin 4 in human lymphocytes preactivated by interleukin 2 in vivo or in vitro.

In an attempt to augment the generation of human cytotoxic effector cells for potential cancer therapy with interleukin 2 (IL2) and lymphokine-activated killer (LAK) cells, the effect of interleukin 4 (IL4) on LAK cell induction was studied. In normal human peripheral blood lymphocytes (PBL), IL4 does not induce LAK activity and inhibits LAK induction by IL2. However, since lymphocyte activation, such as with antigen or mitogen, can render them responsive to IL4, the ability of IL4 to induce LAK activity in lymphocytes preactivated in vivo or in vitro with IL2 was investigated. PBL obtained from 12 patients with advanced cancer 1 to 3 days after IL2 therapy and from eight healthy control subjects were cultured 4 to 5 days with or without IL4 and/or IL2 and then tested for LAK activity as assessed by lysis of Daudi in a 4-h 51Cr release assay. In normal PBL, IL4 failed to induce LAK activity and consistently inhibited LAK induction by a suboptimal concentration of IL2 (10 units/ml). By contrast, IL4 induced LAK activity in PBL from seven of twelve IL2-treated patients and augmented LAK induction by the suboptimal IL2 in PBL from five of twelve IL2-treated patients. With an optimal LAK-inducing concentration of IL2 (1000 units/ml), IL4 less consistently inhibited LAK induction in normal PBL and had a variable effect upon LAK induction in PBL from IL2-treated patients. IL4 induced LAK activity in PBL obtained from a cancer patient after, but not before, systemic IL2 therapy. Similarly, IL4 induced LAK activity in normal PBL only after they had been preincubated with IL2. Thus, IL4 induces LAK activity in lymphocytes preactivated by IL2 in vivo or in vitro. Fluorescence-activated cell sorting revealed that the LAK activity, whether induced by IL4 or by IL2, was mediated largely by non-T (CD5-) natural killer-like (CD56+) cells. The results suggest a regulatory relationship between IL2 and IL4 in the induction and/or maintenance of LAK activity, which might be exploited to augment the generation of cytotoxic cells for lymphokine-mediated immunotherapy of human cancer.

Identification of a unique T cell-derived lymphokine that primes macrophages for tumor cytotoxicity.

Macrophage activation factor (MAF) activity, assessed by the ability to activate macrophages (MO) to lyse RBL--a TNF-resistant, retrovirally transformed, tumor target--was detected in the PHA-stimulated supernatant (Sup) of LBRM, a murine T cell line. LBRM Sup provided a priming signal to MO, but required the subsequent addition of small amounts of LPS for the expression of tumor cytotoxicity. The identity of the lymphokine responsible for this MAF activity was investigated. IFN-gamma, the only previously characterized lymphokine capable of priming MO for tumor cytotoxicity, did have MAF activity in the assay, but IFN-gamma could not be detected by ELISA in LBRM Sup, and LBRM-derived mRNA lacked detectable message for IFN-gamma. Moreover, anti-IFN-gamma failed to inhibit the MAF activity of LBRM Sup, suggesting that the presence of small, undetectable amounts of IFN-gamma were neither responsible nor required for LBRM MAF activity. LBRM MAF activity appeared distinct from the other previously identified lymphokines produced by LBRM, since granulocyte-macrophage-CSF, IL-2, and IL-3 purified from LBRM Sup were unable to activate MO to lyse RBL. IL-4 and TNF, two lymphokines not known to be produced by LBRM but able to activate MO for cytotoxicity of some tumor targets, were also unable to activate MO for RBL cytotoxicity. LBRM MAF lacked antiviral activity in biologic assays, further distinguishing the lymphokine from IFN-gamma, and had an apparent Mr of 30,000 Da using gel filtration chromatography. Thus, the LBRM T cell line produces a previously undescribed lymphokine that primes MO for tumor cytotoxicity.