KRASG12D, pulmonary LCH, and atorvastatin

Achille Aouba

In this issue of Blood, Kamata et al1 elegantly demonstrate in mice that (1) KRASG12D (but not BRAFV600E) mutation transfection in pulmonary myeloid cells induces an isolated pulmonary (IP)-Langerhans cell histiocytosis (LCH)–like neoplasm, via phosphatidylinositol 3-kinase (PI3K) activation, and that (2) atorvastatin, a 3-hydroxy-3-methylglutaryl coenzyme A inhibitor that inhibits RAS interaction of PI3K, reduces tumor burden in vivo.

LCH is a rare and heterogeneous disease that has been called by several names, including “histiocytosis X,” which emphasizes the ambiguity of its cellular origin and pathogenesis. Clinical presentation and prognosis are highly variable. For example, in young adults, the disease often involves a single organ, has a favorable prognosis, and often resolves spontaneously. In infants, the disease presents as a systemic leukemia-like disease that requires chemotherapy. Therefore, age of onset roughly characterizes 2 distinct clinical forms of the disease with distinct etiologies. In adults, 2 additional presentations of the disease are a “systemic” form with single or multiorgan involvement with symptoms such as weight loss and fever or the IP-LCH, in which smoking is a significant risk factor.2 IP-LCH appears, therefore, as a distinct disease with a worse response to treatment than the usual adult systemic disease. The reasons for these differences are not known.

The study by the authors addressed many of these issues by exploring the pathophysiology revealing a new therapeutic option in IP-LCH. Recently, molecular defects have been shown to be involved in both LCH and Erdheim-Chester disease (ECD), a non-Langerhans cell histiocytosis, which share close pathogenic mechanisms and can be found in the same patient and even in the same biopsy site. BRAFV600E and NRAS molecular defects, leading to RAF/RAS/MEK/extracellular signal-regulated kinase activation, have been described in both diseases, leading to their classification as myeloid inflammatory tumors possibly treatable with targeted anti-BRAF or anti-MEK molecules.3,4

The study by Kamata et al clearly shows that IP-LCH and systemic LCH have distinct cellular origins and oncogenic mechanisms in their mouse model. Indeed, the authors demonstrated that IP-LCH originates from pulmonary resident cells that correspond to myeloid and not dendritic cells. Moreover, they showed that only KRASG12D, but not BRAFV600E expression, is able to induce tumorigenesis in these monocytic cells, probably via PI3K pathway activation. These findings separate IP-LCH from systemic LCH that originates in dendritic cells acquiring the BRAF mutation leading to oncoprotein BRAFV6OOE expression, which activates the MEK/extracellular signal-regulated kinase pathway. KRAS mutations have been researched but never described in LCH.3,4 In ECD, only 2 patients with KRAS mutations have been identified, corresponding to a KRASQ61H defect present at baseline in 1 patient and a KRASG12D defect in the context of acquired resistance to an anti-BRAF treatment in the other.3,5 Several authors found concordance among tissue, plasmatic, and/or urinary cell-free DNA for the various mutations involved in ECD or LCH, including the ECD patient with the posttreatment acquired KRASG12D defect.3 Finally, the KRAS mutation could be a specific molecular defect of IP-LCH among all untreated histiocytic disorders; it could even be a hallmark of IP-LCH if other MAPK (MAP2K1 and NRAS) and PI3K pathway (PIK3CA) genes are proven to not induce this single disease location, as shown for BRAF mutation by the study authors. In the literature, supposed IP-LCH cases have been described in young children with secondary extrapulmonary involvement; however, because “systemic” LCH is a relapsing disorder characterized by wide temporal and anatomical distribution of organ involvement over decades, pulmonary-onset LCH is not necessarily IP-LCH. The term “primary pulmonary LCH”6 is a better name than IP-LCH to avoid any potential confusion.

The in vivo demonstration by the authors of IP-LCH–like disease control with atorvastatin is of clinical interest. Indeed, smoking cessation is frequently the only therapeutic intervention in IP-LCH, but actual efficacy of smoking cessation is unknown and patients do not always accurately report their smoking history. Indeed, smoking cessation is sometimes insufficient to control pulmonary function decline, with a need then for steroids and various cytotoxic drugs.2,7 Unlike pediatric LCH for which vinblastine and steroids are the validated initial treatment, a validated first-line chemotherapy regimen is not available for adult LCH. We recently showed in a retrospective study that vinblastine and steroids is an effective and well-tolerated first-line treatment of adult LCH, except in patients with lung involvement and impaired lung function, because of IP-LCH or systemic disease with pulmonary involvement.7 Moreover, it is also not clearly established whether IP-LCH may spread to other organs. Plasmatic and urinary cell-free DNA for KRASG12D defect research should therefore help answer this question.3 Adding atorvastatin to smoking cessation or to other therapeutic approaches in IP-LCH, or even in systemic disease, could be a promising and harmless first-line treatment option.

Kamata et al hypothesized that atorvastatin acts in KRAS mutation-related IP-LCH via the suppression of mevalonate pathway upstream of isoprenoid lipids’ production. These lipids are required for RAS prenylation and function. Because KRAS function is involved in inflammasome and NF-kB production of the proinflammatory interleukin-1b, atorvastatin could also act in IP-LCH by inhibiting this cytokine, which is pathogenic in ECD and possibly LCH.8-10

Even though the authors’ work is exciting, there are 2 concerns: mouse models are not always comparable to humans and hope raised by the use of 3-hydroxy-3-methylglutaryl coenzyme A-reductase inhibitors in the treatment of the inflammatory component of mevalonate kinase deficiency have been rather disappointing so far. However, considering the widespread use and the relative safety of these drugs in clinical practice, their assessment in IP-LCH should be undertaken rapidly.


  • Conflict-of-interest disclosure: The author declares no competing financial interests.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.