The Elephant Man Meets Pulmonary Hypertension. A Cautionary Tale.

Neurofibromatosis (NF) has achieved notoriety because of Joseph Merrick, a medical and sideshow phenomenon in the late 1800s in London who was diagnosed with NF in 1909 (1). His life has been chronicled in several books and films, including the critically acclaimed film The Elephant Man in 1980, as well as theatrical productions in both London and New York City. From these, NF became more accepted and investigated (found to be three subtypes: NF1, NF2, and schwannomatosis), and the genetic mutations have been identified (2). Over the years, complications and issues associated with the neurofibromatoses have become apparent. In this issue of the Journal, Jutant and colleagues (pp. 843–852) describe a little-appreciated aspect of NF1, pulmonary hypertension (PH) (3).

PH is a rare and incompletely characterized complication of NF1. First described in 1986, the largest previously reported series included just eight patients and was notable for a poor response to PH-specific therapy and poor outcomes (4, 5). Since that report in 2011, individual cases of PH-NF1 have appeared in the literature. In this issue of the Journal, Jutant and colleagues, using data from the French Pulmonary Hypertension Network, describe clinical, functional, hemodynamic, and radiographic characteristics as well as responses to pulmonary arterial hypertension (PAH)-specific therapy in 49 cases of PH-NF1, thereby comprising the largest and most comprehensive series to date; in fact, this series is greater than the total number of cases of PH-NF1 reported thus far. Though largely confirming many of the smaller previous reports, what emerges from this study not only paints a foreboding picture of PH-NF1 but also raises many additional questions.

PH-NF1 is largely a late complication of NF1 with a median age of diagnosis of 62 years. Interestingly, there is a nearly 4:1 female predominance, in keeping with the female predominance noted in idiopathic and heritable PAH and raising the specter of hormonal influence on disease development (6, 7). Patients largely presented with advanced disease at diagnosis, with New York Heart Association functional class III or IV, 6-minute-walk distances <250 m, and severe precapillary PH by hemodynamics with a mean pulmonary artery pressure of 45 mm Hg, a pulmonary vascular resistance of 10.7 WU, and cardiac index of 2.3 L/min/m2. Patients with PH-NF1 were poorly responsive to therapy with high mortality (46% 5-yr survival), even in the setting of combination pulmonary vasodilator therapy, including intravenous prostacyclin in some cases.

The poor outcomes and response to therapy that have been previously reported and again confirmed by Jutant and colleagues may be multifactorial and related, in part, to the phenotypic complexity of PH-NF1. Having been associated with vascular remodeling, interstitial lung disease, left heart disease, and skeletal abnormalities leading to secondary restrictive cardiopulmonary physiology, PH-NF1 is quite deservedly classified as World Health Organization group 5 PH, PH secondary to unclear/multifactorial mechanisms (5, 8, 9). Indeed, in the cohort presented by Jutant and colleagues, pulmonary parenchymal involvement was noted in 40 of the 41 patients with interpretable high-resolution computed tomography scans. This observation, in combination with the frequent hypoxemia noted at diagnosis in the cohort, suggests a prominent role in the pathophysiology from parenchymal lung involvement; yet, 27 patients had normal spirometry and lung volumes. Interestingly, the predominant pulmonary function test abnormality was a severely reduced diffusion capacity of the lung for carbon monoxide, likely speaking to the significance of pulmonary vascular involvement in the overall phenotype but certainly not excluding some form of parenchymal involvement.

More interesting still, of the three available pathologic samples in the cohort, parenchymal abnormalities as well as severe arterial and venous remodeling were noted in all samples, the latter most concerning for a pulmonary venoocclusive disease like pathophysiology with subsequent implications for poor response to PAH-specific therapy (10, 11). Though it is unlikely that three samples are fully representative of the pathologic spectrum of PH-NF1, these findings corroborate what has been previously reported in other series and, taken together, speak to the phenotypic heterogeneity present in PH-NF1 (4, 9).

Because of the functional and hemodynamic severity of the PH in this cohort, the physicians caring for them ultimately opted to treat 45 of the 49 patients with PAH-specific medications, including 44% treated with combination pulmonary vasodilator therapy on second follow-up and 64% on the last reassessment. With this modern-era PAH treatment regimen, despite improvements in hemodynamics and New York Heart Association functional class, hypoxemia worsened irrespective of the severity of spirometry or lung volume abnormalities by pulmonary function testing, and 6-minute-walk distance remained unchanged initially but decreased below baseline at the last reassessment. Even more concerning, however, is that during the course of treatment with PAH-specific therapy, three patients died suddenly at home of unclear causes, and overall mortality was much higher in the PH-NF1 cohort compared with the their idiopathic PAH counterparts (12). Thus, despite some evidence of short-term benefit, routine treatment of these patients with currently available PAH-specific medications cannot be recommended. Rather, based on available observations, it seems more prudent to focus on nonspecific treatment with oxygen and diuretics as indicated with early referral for lung transplant in those who are eligible.

Although we congratulate the authors for the most complete description to date of PH-NF1 and the largest cohort presently available, it is still a relatively small, retrospective sample. Despite this, it seems that PH remains a rare, phenotypically heterogeneous complication of NF1 with poor outcomes and no conclusive data to support treatment with pulmonary vasodilators. Increased awareness of PH-NF1 among providers with a low threshold for screening echocardiogram, high-resolution computed tomography, and right heart catheterization when indicated among symptomatic patients is imperative to target earlier diagnosis and may help amass larger cohorts that can be studied prospectively to devise treatment regimens to improve short- and long-term outcomes.

As for Mr. Merrick, we bet you thought we were going to tell you that he died from what appeared to be PH; alas, no—he died from asphyxiation. Moreover, although his physical condition was long attributed to NF1, some researchers believe that he may actually have suffered from the even rarer Proteus syndrome; however, despite genetic analysis of his hair and bone in 2003, the exact etiology of his deformities has never been conclusively established (13).