Early data on long-term efficacy and safety of inotersen in patients with hereditary transthyretin amyloidosis: a 2-year update from the open-label extension of the NEURO-TTR trial.

BACKGROUND AND PURPOSE: Hereditary transthyretin (hATTR) amyloidosis causes progressive polyneuropathy resulting from transthyretin (TTR) amyloid deposition throughout the body, including the peripheral nerves. The efficacy and safety of inotersen, an antisense oligonucleotide inhibitor of TTR protein production, were demonstrated in the pivotal NEURO-TTR study in patients with hATTR polyneuropathy. Here, the long-term efficacy and safety of inotersen are assessed in an ongoing open-label extension (OLE) study. METHODS: Patients who completed NEURO-TTR were eligible to enroll in the OLE (NCT02175004). Efficacy assessments included the modified Neuropathy Impairment Score plus seven neurophysiological tests composite score (mNIS + 7), the Norfolk Quality of Life - Diabetic Neuropathy (Norfolk QOL-DN) questionnaire total score and the Short-Form 36 Health Survey (SF-36) Physical Component Summary (PCS) score. Safety and tolerability were also assessed. RESULTS: Overall, 97% (135/139) of patients who completed NEURO-TTR enrolled in the OLE. Patients who received inotersen for 39 cumulative months in NEURO-TTR and the OLE continued to show benefit; patients who switched from placebo to inotersen in the OLE demonstrated improvement or stabilization of neurological disease progression by mNIS + 7, Norfolk QOL-DN and SF-36 PCS. No new safety concerns were identified. There was no evidence of increased risk for grade 4 thrombocytopenia or severe renal events with increased duration of inotersen exposure. CONCLUSION: Inotersen slowed disease progression and reduced deterioration of quality of life in patients with hATTR polyneuropathy. Early treatment with inotersen resulted in greater long-term disease stabilization than delayed initiation. Routine platelet and renal safety monitoring were effective; no new safety signals were observed.

Retrospective study of a TTR FAP cohort to modify NIS+7 for therapeutic trials.

Protein stabilization and oligonucleotide therapies are being tested in transthyretin amyloid polyneuropathy (TTR FAP) trials. From retrospective analysis of 97 untreated TTR FAP patients, we test the adequacy of Neuropathy Impairment Score+7 tests (NIS+7) and modifications to comprehensively score impairments for use in such therapeutic trials. Our data confirms that TTR FAP usually is a sensorimotor polyneuropathy with autonomic features which usually is symmetric, length dependent, lower limb predominant and progressive. NIS+7 adequately assesses weakness and muscle stretch reflexes without ceiling effects but not sensation loss, autonomic dysfunction or nerve conduction abnormalities. Three modifications of NIS+7 are suggested: 1) use of Smart Somatotopic Quantitative Sensation Testing (S ST QSTing); 2) choice of new autonomic assessments, e.g., sudomotor testing of distributed anatomical sites; and 3) use of only compound muscle action potential amplitudes (of ulnar, peroneal and tibial nerves) and sensory nerve action potentials of ulnar and sural nerve - than the previously recommended attributes suggested for the sensitive detection of diabetic sensorimotor polyneuropathy. These modifications of NIS+7 if used in therapeutic trials should improve characterization and quantification of sensation and autonomic impairment in TTR FAP and provide better nerve conduction tests.

Safety and efficacy of high-dose melphalan and auto-SCT in patients with AL amyloidosis and cardiac involvement.

In Ig light chain (AL) amyloidosis, cardiac involvement is associated with worse prognosis and increased treatment-related complications. In this retrospective cohort study, we assessed survival, hematologic and cardiac responses to high-dose melphalan and auto-SCT (HDM/SCT) in patients with AL amyloidosis and cardiac involvement, stratified by cardiac biomarkers brain natriuretic peptide and Troponin I, analogous to the Mayo cardiac staging. Forty-seven patients underwent HDM/SCT based upon functional measures; six patients had modified cardiac stage I disease, seventeen had modified cardiac stage II disease and twenty-four had modified cardiac stage III disease. Treatment-related mortality was 4% for all patients and 8% for patients with stage III disease. Three-year survival was 88% and EFS was 47%; these did not differ by stage. By intention-to-treat analysis, 27% of patients achieved a hematologic complete response and 32% a very good partial response, of whom 70 and 45%, respectively, have not required additional therapy at 36 months. Cardiac response was achieved in 53% of patients. We conclude that with appropriate patient selection and a risk-adapted treatment approach, HDM/SCT is safe and effective in patients with AL amyloidosis and cardiac involvement.

Tandem cycles of high-dose melphalan and autologous stem cell transplantation increases the response rate in AL amyloidosis.

Clinical outcomes of patients with AL amyloidosis treated with high-dose melphalan and stem cell transplantation (HDM/SCT) are tightly linked to the achievement of a hematologic complete response (HCR). We conducted a prospective trial to determine whether a second cycle of HDM/SCT could induce HCR in patients in whom the plasma cell dyscrasia persisted following initial treatment with HDM/SCT. Sixty-two patients were enrolled. Nine patients (15%) were removed from the protocol. Of the 53 patients continuing in this study, four died within 100 days of treatment (8%), and 27 (55%) achieved an HCR at 6 months after the first cycle of HDM/SCT. Of the 22 patients who did not achieve an HCR after initial treatment, 17 received a second HDM/SCT, 1 died within 100 days of treatment (6%), while 5 (31%) achieved an HCR. Thus, the HCR rate was 67% (32/48) for surviving patients on study, 60% (32/53) for all patients who received initial cycle of HDM/SCT, and 56% (35/62) by intention-to-treat. The median survival for all patients enrolled on the trial has not yet been reached. Thus, tandem cycles of HDM/SCT can increase the proportion of patients who achieve an HCR.

High-dose intravenous melphalan and autologous stem cell transplantation as initial therapy or following two cycles of oral chemotherapy for the treatment of AL amyloidosis: results of a prospective randomized trial.

SUMMARY: A prospective randomized trial was conducted to study the timing of high-dose intravenous melphalan and autologous stem cell transplantation (HDM/SCT) in AL amyloidosis. In all, 100 newly diagnosed patients were randomized to receive HDM/SCT, either as initial therapy (Arm-1) or following two cycles of oral melphalan and prednisone (Arm-2). The objectives of the trial were to compare survival and hematologic and clinical responses. With a median follow-up of 45 months (range 24-70), the overall survival was not significantly different between the two treatment arms (P=0.39). The hematologic response and organ system improvements after treatment did not differ between the two groups. Fewer patients received HDM/SCT in Arm-2 because of disease progression during the oral chemotherapy phase of the study, rendering them ineligible for subsequent high-dose therapy. This affected patients with cardiac involvement particularly, and led to a trend for an early survival disadvantage in Arm-2. Hence, newly diagnosed patients with AL amyloidosis eligible for HDM/SCT did not benefit from initial treatment with oral melphalan and prednisone, and there was a survival disadvantage for patients with cardiac involvement if HDM/SCT was delayed by initial oral chemotherapy.

Effect of dose-intensive intravenous melphalan and autologous blood stem-cell transplantation on al amyloidosis-associated renal disease.

BACKGROUND: Dose-intensive intravenous melphalan with autologous blood stem-cell transplantation induces remission of the plasma cell dyscrasia in a substantial proportion of patients with AL amyloidosis. The impact of this treatment on associated renal disease is not known. OBJECTIVE: To determine the effect of dose-intensive intravenous melphalan and autologous blood stem-cell transplantation on AL amyloidosis-associated renal disease. DESIGN: Prospective cohort study. SETTING: Academic medical center. PATIENTS: 65 patients with AL amyloidosis and urinary protein excretion greater than 1 g/24 h who received dose-intensive intravenous melphalan and autologous blood stem-cell transplantation between 1 July 1994 and 30 June 1998. MEASUREMENTS: 24-hour urinary protein excretion, serum cholesterol level, serum albumin level, creatinine clearance, urine and serum immunoelectrophoresis, and bone marrow biopsy. Renal response was defined as a greater than 50% reduction in urinary protein excretion in the absence of a 25% or greater reduction in creatinine clearance. Complete hematologic response was defined as absence of detectable monoclonal protein in serum and urine and a bone marrow specimen containing less than 5% plasma cells without clonal dominance of kappa or lambda isotype. RESULTS: Among the 50 patients who survived for at least 12 months, proteinuria, hypoalbuminemia, and hypercholesterolemia improved during follow-up; 36% met criteria for a renal response. Median 24-hour urinary protein excretion decreased from a baseline value of 9.6 g/24 h to 1.6 g/24 h at 12 months among patients with complete hematologic response, and 71% met criteria for a renal response. Twenty-hour urinary protein excretion did not decrease during follow-up among patients with persistent plasma cell disease, and only 11% had a renal response at 12 months (P < 0.001 for hematologic responders vs. nonresponders). CONCLUSION: Dose-intensive intravenous melphalan with autologous blood stem-cell transplantation improves the nephrotic syndrome in patients with AL amyloidosis-associated renal disease. The benefit is largely limited to patients achieving eradication of the underlying plasma cell dyscrasia.

Hypoxia inhibits amino acid uptake in human lung fibroblasts.

Hypoxia and amino acid deprivation downregulate expression of extracellular matrix genes in lung fibroblasts. We examined the effect of hypoxia on amino acid uptake and protein formation in human lung fibroblasts. Low O(2) tension (0% O(2)) suppressed incorporation of [(3)H]proline into type I collagen without affecting [(35)S]methionine labeling of other proteins. Initial decreases in intracellular [(3)H]proline incorporation occurred after 2 h of exposure to 0% O(2), with maximal suppression of intracellular [(3)H]proline levels at 6 h of treatment. Hypoxia significantly inhibited the uptake of radiolabeled proline, 2-aminoisobutyric acid (AIB), and 2-(methylamino)isobutyric acid (methyl-AIB) while inducing minor decreases in leucine transport. Neither cycloheximide nor indomethacin abrogated hypoxia-related suppression of methyl-AIB uptake. Efflux studies demonstrated that hypoxia inhibited methyl-AIB transport in a bidirectional fashion. The downregulation of amino acid transport was not due to a toxic effect; function recovered on return to standard O(2) conditions. Kinetic analysis of AIB transport revealed a 10-fold increase in K(m) accompanied by a small increase in maximal transport velocity among cells exposed to 0% O(2). These data indicate that low O(2) tension regulates the system A transporter by decreasing transporter substrate affinity.

Tracheobronchial amyloidosis. The Boston University experience from 1984 to 1999.

Tracheobronchial amyloidosis (TBA), an idiopathic disorder characterized by deposition of fibrillar proteins in the tracheobronchial tree, occurred in 10 patients referred to the Amyloid Program at Boston University over the past 15 years. Fewer than 100 cases of TBA have been described; only 1 series encompassed more than 3 patients. We analyzed our experience with biopsy-proven TBA to define better its natural history. Follow-up averaged approximately 8 years and was obtained in all cases, making this outcome reporting the largest and most complete to date. Three of these patients were prospectively studied for up to 24 months to examine the utility of bronchoscopy, computerized tomography (CT) imaging, and pulmonary function tests (PFTs) in monitoring disease progression. No patient with TBA developed signs or symptoms of systemic amyloidosis during the period reviewed. Conversely, tracheobronchial disease was not diagnosed in 685 patients with primary systemic (AL) amyloidosis during the 15-year study period at Boston University. Bronchoscopy proved most useful in establishing the diagnosis by biopsy. Narrowing of major airways limited its inspection of the tracheobronchial tree, however. In contrast, CT imaging provided quantitative assessment of airway narrowing and mural thickening--2 major consequences of amyloid infiltration. These CT features, in the presence of mural calcifications sparing the posterior tracheal membrane, have been reported in few disorders other than TBA. The ability of CT to map airway involvement and identify extraluminal manifestations of TBA made it the study of choice for establishing disease extent. Three patterns of disease were evident by CT imaging and bronchoscopic examination: proximal, mid, and distal airways involvement. Those with severe proximal disease had significantly decreased air flows, air trapping, and fixed upper airway obstruction on PFTs. Patients with distal disease had normal airflows. PFTs could not clearly distinguish proximal from severe mid airways disease. Thirty percent of patients died within 7-12 years after diagnosis, all having proximal or severe mid airways disease. Repeated rigid bronchoscopic debridement and laser treatments did not prevent progressive airways narrowing in patients dying from TBA. Most patients with mid airways involvement, and all distal airway cases, had either stagnant disease or slowly increasing amyloid deposits when followed for up to 14 years. In a small subset of patients followed prospectively, serial PFTs were most sensitive to disease progression. CT-derived measures of airway lumen diameter and wall thickness did not change significantly despite marked improvements in airflow after rigid bronchoscopy. Our experience suggests that serial PFTs and CT imaging together offer the best assessment of airway involvement and disease progression in patients with TBA. In the future, radiation therapy may provide more definitive treatment of TBA than debulking procedure have to date.

Hypoxia downregulates tropoelastin gene expression in rat lung fibroblasts by pretranslational mechanisms.

Elastolytic lung injury disrupts cell barriers, flooding alveoli and producing regional hypoxia. Abnormal O2 tensions may alter repair of damaged elastin fibers. To determine the effect of hypoxia on extravascular elastin formation, we isolated rat lung fibroblasts and cultured them under a variety of O2 conditions. Hypoxia downregulated tropoelastin mRNA in a dose- and time-related fashion while upregulating glyceraldehyde-3-phosphate dehydrogenase mRNA levels. The changes in tropoelastin gene expression were not due to cell toxicity as measured by chromium release and cell proliferation studies. Neither cycloheximide nor actinomycin D abrogated this effect. Hypoxia induced early decreases in tropoelastin mRNA stability; minor suppression of gene transcription occurred later. When returned to 21% O2, tropoelastin mRNA recovered to control levels in part by upregulating tropoelastin gene transcription. Taken together, these data indicate that hypoxia regulates tropoelastin gene expression and may alter repair of acutely injured lung.

Intermediate-dose intravenous melphalan and blood stem cells mobilized with sequential GM+G-CSF or G-CSF alone to treat AL (amyloid light chain) amyloidosis.

AL amyloidosis patients ineligible for dose-intensive melphalan (200 mg/m2) were enrolled on a phase 11 trial to be treated with two cycles of intermediate-dose melphalan (IDM 100 mg/m2) and mobilized blood stem cells (BSC). For mobilization patients were randomized to either GM-CSF 250 microg/m2 for 3 d followed by G-CSF 10 microg/ kg for 3 d (GM+G), or G-CSF 10 microg/kg for 6 d (G-alone), with leukaphereses on days 5, 6 and 7. To minimize morbidity, we planned to support each cycle with 3 5 x 106 CD34+ cells/kg and had a collection target of 7 x 10(6) CD34+ cells/kg. Those who did not achieve the target were treated with one cycle of IDM. 30 patients, a median of 62 years old and 7 months from diagnosis, were enrolled. Both mobilization regimens were generally well tolerated, and similar in terms of CD34+ cells and CFU-GM collected, but only 6/28 patients achieved the collection target (GM+G four, G-alone two). Despite a 19% incidence of grade 4 toxicities, IDM therapy was well tolerated. At a median follow-up of 24 months (19-36) 57% of patients had survived, 17% with durable complete haematological responses and 40% with improved or stable amyloid organ involvement, including 3/9 patients with predominant cardiac amyloid who are alive 2-3 years after treatment. The 100 d mortality was 20%. In conclusion, no definitive differences were identified between the mobilization regimens and IDM was an active regimen in AL for selected patients.

Dose-intensive melphalan with blood stem-cell support for the treatment of AL (amyloid light-chain) amyloidosis: survival and responses in 25 patients.

AL (amyloid light-chain) amyloidosis is an uncommon plasma cell disorder in which depositions of amyloid light-chain protein cause progressive organ failure and death in a median of 13 months. Autologous stem-cell transplantation is effective therapy for multiple myeloma and therefore, we evaluated its efficacy for AL amyloidosis. Patients with adequate cardiac, pulmonary, and renal function had stem cells mobilized with granulocyte-colony stimulating factor and were treated with dose-intensive intravenous melphalan (200 mg/m2). Response to therapy was determined by survival and improvement of performance status, complete response or persistence of the clonal plasma cell disorder, and change in the function of organs involved with amyloid at baseline. We enrolled 25 patients with a median age of 48 years (range, 29-60), all of whom had biopsy-proven amyloidosis with clonal plasma cell disorders. Twenty-two (88%) were Southwest Oncology Group performance status 1 or 2 within a year of diagnosis, and 16 (64%) had received no prior therapy. Predominant amyloid-related organ involvement was cardiac (n = 8), renal (n = 7), hepatic (n = 6), neuropathic (n = 3), and lymphatic (n = 1). Fifteen patients had one or two organ systems involved, whereas 10 had three or more involved. With a median follow-up of 24 months (12-38), 17 of 25 patients (68%) are alive, and the median survival has not been reached. Thirteen of 21 patients (62%) evaluated 3 months posttransplant had complete responses of their clonal plasma cell disorders. Currently, two thirds of the surviving patients (11 of 17) have experienced improvements of amyloid-related organ involvement in all systems, whereas 4 of 17 have stable disease. The improvement in the median performance status of the 17 survivors at follow-up (0 [range, 0-3]) is statistically significant versus baseline (2 [range, 1-3]; P < . 01). Significant negative prognostic factors with respect to overall survival include amyloid involvement of more than two major organ systems and predominant cardiac involvement. Three patients have experienced relapses of the clonal plasma cell disorder at 12 and 24 months. Dose-intensive therapy should currently be considered as the preferred therapy for patients with AL amyloidosis who meet functional criteria for autologous transplantation.

Effect of okadaic acid on elastin gene expression in interstitial lung fibroblasts.

Okadaic acid (OA), a specific serine/threonine protein phosphatase inhibitor, downregulated tropoelastin formation and elastin mRNA levels in a dose-related and cycloheximide-sensitive fashion in cultured lung fibroblasts. Treatment with a tyrosine phosphatase inhibitor at high concentrations did not alter elastin mRNA levels, however. Nuclear run-on analysis indicated that OA primarily suppressed elastin gene expression through a transcriptional mechanism. In contrast to its effects on elastin expression, OA downregulated alpha 1(I) mRNA to significantly lesser degrees. The mechanism by which OA decreased elastin mRNA levels did not appear to involve protein kinase C or share the signaling pathway of IL-1 beta. Prolonged treatment with phorbol ester promoted the inhibitory effects of OA on elastin, as did shorter treatment with IL-1 beta. Moreover, transient transfection studies indicated that OA and IL-1 beta do not act through the same cis-acting element in the elastin promoter. Finally, unlike the transient effects of IL-1 beta, OA induced persistent inhibition of elastin expression by a transcriptional mechanism. Taken together, these data indicate that serine/threonine protein phosphorylation can regulate the amount and composition of extracellular matrix secreted by fibroblasts into the interstitium of the lung.

Regulation of type I collagen production by insulin and transforming growth factor-beta in human lung fibroblasts.

The effects and interaction of transforming growth factor-beta (TGF-beta) and insulin on collagen production in human fetal lung fibroblasts was examined. Fibroblasts were labeled with [3H]proline and collagen production was analyzed by polyacrylamide gel electrophoresis. The addition of insulin (2 micrograms/ml) increased collagen production 5 fold and TGF-beta (5 ng/ml) increased collagen production 6-fold. The combination of TGF-beta and insulin further increased type I collagen production (12 fold). We found that TGF-beta increased pro-alpha 1 (I) collagen mRNA levels 2-3 fold, insulin increased mRNA levels by less than 2 fold, and the combination stimulated a 3-4 fold increase. In a nuclear run-on assay, we found a 1.7 fold increase in the rate of transcription for the pro-alpha 1 (I) collagen gene in insulin-treated cultures and a 2-fold increase in TGF-treated cultures. In fibroblasts transfected with a plasmid containing 2.4 kb of the 5' flanking sequences of the human pro-alpha 1 (I) collagen gene, TGF-beta stimulated a 2.8 fold increase in promoter activity. In contrast, the addition of insulin stimulated a small increase (less than 2 fold) in the pro-alpha 1 (I) collagen promoter activity when administered alone or in combination with TGF-beta. Insulin prolonged the half-life of pro-alpha 1 (I) collagen mRNA from 9.1 h to 14.3 h as assessed by treatment with actinomycin D. The insulin-induced increase in pro-alpha 1 (I) collagen mRNA was blocked by the presence of cycloheximide indicating a requirement for new protein synthesis. These results show that the combination of TGF-beta and insulin stimulate large increases in type I collagen formation by acting at different sites in the collagen biosynthetic pathway.

Retinoic acid-induced inhibition of type I collagen gene expression by human lung fibroblasts.

We examined alpha 1(I) collagen gene expression in all-trans retinoic acid (RA)-treated human lung fibroblast cultures. RA (10(-5)M) decreased steady-state levels for alpha 1(I) collagen mRNA by at least 75% after 24 h. The inhibition was evident within 8 h after addition of RA, was dose dependent and reversible. Treatment with 9-cis-retinoic acid did not affect alpha 1(I) collagen mRNA levels. RA also inhibited the increases in alpha 1(I) mRNA stimulated by transforming growth factor-beta (TGF-beta). The RA-mediated decrease in alpha 1(I) collagen mRNA was blocked by cycloheximide treatment, suggesting that synthesis of a protein intermediate is required for the inhibition. The RA-induced decrease in alpha 1(I) collagen mRNA levels was not mediated by increases in prostaglandin E2 production. RA decreased alpha 1(I) gene transcription as determined by nuclear run-off assays but did not significantly alter the rate of degradation of the alpha 1(I) transcript as determined by actinomycin D treatment. Studies employing cells stably transfected with constructs containing portions of the alpha 1(I) collagen promoter indicate that the DNA sequences which mediate the inhibitory effect are located within 900 bases from the transcription start site.

The effect of retinoic acid on amino acid uptake and protein synthesis by lung fibroblasts.

The effect of retinoic acid (RA) on the uptake and utilization of extracellular amino acids by fetal lung fibroblasts was examined. RA decreased the incorporation of [3H]proline into collagen and other proteins. The effect was maximal at a RA concentration of 10(-5) M; smaller decreases were observed at a RA concentration of 10(-6) M. This decrease in collagen formation was associated with a large decrease in intracellular [3H] proline. The decrease in intracellular [3H]proline was first observed at 2 h following the addition of RA to cell cultures. Transport studies employing radiolabeled amino acids revealed that RA decreased the uptake of proline, 2-aminoisobutyric acid, and 2-(methylamino)isobutyric acid but not leucine or methionine. Kinetic analysis of 2-aminoisobutyric acid uptake indicated that this effect was mediated primarily by an increase in apparent Km, with a lesser decrease in Vmax, RA-induced inhibition of proline uptake was not abolished by the presence of cycloheximide nor by pretreatment with indomethacin. Na+,K(+)-ATPase activity was not affected by RA treatment. These results suggest that RA modulates protein production in fibroblasts by altering the function of the Na(+)-dependent A transport system for amino acid uptake.

Recombinant interleukin-1 beta inhibits elastin formation by a neonatal rat lung fibroblast subtype.

The effect of recombinant interleukin-1 beta (rIL-1 beta) on elastin accumulation by lipid-laden interstitial cells (LIC) derived from neonatal rat lung was examined. The LIC, a fibroblast subtype, synthesized large amounts of elastin which was deposited into the extracellular matrix. This elastin was alkali-resistant and had an amino acid composition typical of adult rat elastin. Treatment of lipid-laden interstitial cell cultures with rIL-1 beta at 100 pg/ml caused a dramatic decrease in elastin accumulation as assessed by hot alkali treatment and transmission electron micrographs of the cell cultures. Tropoelastin formation was selectively decreased by rIL-1 beta relative to other proteins. Steady state levels of elastin mRNA were slightly decreased by rIL-1 beta at 5 pg/ml and markedly decreased by rIL-1 beta at 50 pg/ml or greater. The addition of indomethacin had no effect on rIL-1 beta-induced decreases in elastin mRNA levels. Inhibiting protein synthesis with cycloheximide blocked the effect of rIL-1 beta on elastin mRNA levels. The level of alpha 1(I) collagen mRNA was decreased by rIL-1 beta, but only at concentrations higher than that needed to induce a decrease in elastin mRNA. These data indicate that rIL-1 beta decreased steady state levels for elastin mRNA and elastin accumulation and can selectively regulate the accumulation of elastin and collagen.

Cold-induced bronchoconstriction: role of cutaneous reflexes vs. direct airway effects.

To determine the relative contributions of direct airway vs. reflex cutaneous thermal receptor stimulation in cold-induced bronchoconstriction, we isolated these two aspects of cold exposure in 10 asthmatics and 13 normal subjects. Ice packs were applied to the skin of the face, chest, thigh, and upper arm in random sequence while serially measuring specific conductance. In this fashion a limited mapping of skin-mediated bronchoconstriction was established. Warm packs were applied to the same areas of control for any potential nonspecific stimulatory effects. Cooling the skin induced bronchoconstriction to a similar degree in both groups; this effect was very small, did not induce symptoms, and was only seen with stimulation of the face. At another time, the subjects performed isocapnic hyperventilation of frigid air to ascertain their response to direct airway cooling. A moderate but significant correlation existed between skin and airway sensitivity; however, the magnitude of the two responses differed markedly. Breathing cold air at rest had no effect on lung function; however, elevating ventilation promptly produced bronchial narrowing. Hence, in a cold environment, the most potent stimulus for the development of airway obstruction in asthmatics derives from a direct airway effect.